Advances in High-speed Rail Technology

Limin Jia
Xuelei Meng
Yong Qin

Train Operation
in Emergencies
Advances in High-speed Rail Technology
More information about this series at http://www.springer.com/series/13506
Limin Jia Xuelei Meng Yong Qin
• •

Train Operation
in Emergencies

123
Limin Jia Yong Qin
Beijing Jiaotong University Beijing Jiaotong University
Haidian District, Beijing Haidian District, Beijing
China China

Xuelei Meng
Lanzhou Jiaotong University
Anning District, Lanzhou, Gansu
China

ISSN 2363-5010 ISSN 2363-5029 (electronic)

Advances in High-speed Rail Technology
ISBN 978-981-10-4596-7 ISBN 978-981-10-4597-4 (eBook)
DOI 10.1007/978-981-10-4597-4
Library of Congress Control Number: 2017937504

© Springer Nature Singapore Pte Ltd. 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore
Preface

Railway transportation, as an important public transport form, has its constant aim,
which is to transport passengers and freights safely, rapidly, reliably, punctually,
and economically. With the increase in the mileage of the Chinese railway network,
the requirement of higher, faster, and more for railway transportation is proposed.
However, natural disasters affecting railway are characterized in recent years by
universality, frequency, and variety. Moreover, there are occasional railway acci-
dents railway safety public incidents, which declined the capacity of the railway
line and reduced the safety and efficiency of the passenger and freight transporta-
tion. Therefore, research on an objective basis to study the train operating problem
in emergencies is required.
On the other hand, with the increase of available mileage in railway (especially
high-speed railway), the topology structure of the railway network is changing
profoundly. A new railway network is forming gradually, which provides condi-
tions for organizing the train operating work based on the network and makes the
train operating work more complicated. It is urgent to study the train operating
problem based on the railway network.
Based on this background, this book focuses on the theories and methods for
train operation organization in emergencies, using the top-down system analysis
method.
The research work includes railway transport organization mode in emergencies,
railway service network reconstruction, capacity calculation in emergencies, gen-
eration of train paths, train repathing and train re-scheduling problem. The railway
transportation mode in emergencies is the most macro, the most fundamental issue,
which proposes the most basic constraints for train operation organization. It is a
strategic level problem.
And the train service network is the intuitive representation of the line plan,
deciding the origin station, destination station, train path, stops plan, service fre-
quency, and other factors, which is a tactical level problem.
Capacity calculation is the basis of train repathing problem, which is also studied
in this book.

v
vi Preface

At the operational level, the train timetable on the dispatching sections is the key
issue, determining the inbound and outbound time of the trains at stations.
The solution of strategic level problem is the constraint of the tactical level
problem, and the solution of the tactical level problem is the constraint of the
operational level problem. Conversely, the solution of the operational level problem
can be fed back to the tactical level problem, which can be helpful for the
decision-makers of the tactical level problem. In addition, the solution of the tactical
level problem can be fed back to the strategic level problem, which can generate
some decision-supporting information for the strategic level problem.
The authors and their team are involved in the research work of train operation
organization for many years. They have undertaken and completed China National
Key Research and Development Project (Grant: 2016YFB1200105). The authors
developed a series of computer management systems and obtained the software
copyright of train operation regulation system for high-speed railway network. The
research work of these projects and the results of the work provide materials for this
book.
Here the authors want to express thanks to Vice Professor Xu Jie and Li Wang
for their valuable advice in writing this book.
When writing this book, the authors searched for various publications. We tried
to keep a style of clear definition, with lucid brain, so as to make all kinds of readers
have a clear understanding of the transportation organization and train operation in
emergencies. Due to the author’s knowledge level and the depth and breadth of the
study, the views, methods, and theories mentioned in the book certainly have some
deficiencies. Do not hesitate connecting with the authors to provide your priceless
advice.

Beijing, China Limin Jia

Lanzhou, China Xuelei Meng
Beijing, China Yong Qin
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Introduction on Train Operation Problem in Emergencies . . . . . . . 1
1.1.1 Statement of the Problem . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.2 Systems on Railway Train Operation Dispatching . . . . . . . 2
1.1.3 Future Railway Operation Dispatching Mode . . . . . . . . . . . 3
1.1.4 Significance and Necessity of This Research . . . . . . . . . . . 4
1.2 Related Works Review and Analysis . . . . . . . . . . . . . . . . . . . . . . . 5
1.2.1 Related Works on This Problem . . . . . . . . . . . . . . . . . . . . . 5
1.2.2 Analysis on the Publications . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3 Structure of This Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2 Theories on Train Operation in Emergencies . . . . . . . . . . . . . . . . . . . 19
2.1 Introduction on Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.1.1 Definition and Classification of Railway Emergency
Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.1.2 Characteristics of Railway Emergency Events . . . . . . . . . . . 23
2.2 Railway Transport System in Emergencies . . . . . . . . . . . . . . . . . . . 25
2.2.1 System Type of Railway Transportation System . . . . . . . . . 25
2.2.2 System Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2.3 Boundary of the System . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.4 Function and Behavior of the System . . . . . . . . . . . . . . . . . 27
2.2.5 Evolution and Reconstruction of the System . . . . . . . . . . . . 28
2.3 Macroscopic Model of Railway Transport Organization
in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.3.1 General Principles and Methods of Emergency Handling
in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.3.2 Macroscopic Model of Railway Transport Organization
in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.4 Difference Between Train Operation Organization and Normal
Condition in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

vii
viii Contents

2.4.1 Different Goals of Transport Organization . . . . . . . . . . . . . 34

2.4.2 Different Orders of Operation Plan . . . . . . . . . . . . . . . . . . . 36
2.5 Train Operation Organization Strategy in Emergencies. . . . . . . . . . 36
2.5.1 Connotation of Train Operation Organization
in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.5.2 Basic View of Train Operation Organization
in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.5.3 Basic Method of Train Operation Organization
in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.5.4 Principles and Measures of Train Operation Organization
in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3 Transport Organization Mode in Emergencies . . . . . . . . . . . . . . . . . . 43
3.1 Transportation Organization Modes . . . . . . . . . . . . . . . . . . . . . . . . 43
3.1.1 Different Transportation Organization Modes
of High-Speed Railway of Several Representative
Countries and Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.1.2 Analysis of Different Kinds of Transportation Modes . . . . . 47
3.2 Transport Organization Modes in Emergencies . . . . . . . . . . . . . . . . 48
3.2.1 The Necessity of Adjusting Transportation Organization
Mode in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.2.2 The Feasibility of Adjusting the Transportation
Organization Mode in Emergencies. . . . . . . . . . . . . . . . . . . 49
3.2.3 The Model of Transportation Organization Mode
Selection in Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4 Calculation of Railway Transport Capacity in an Emergency
Based on Markov Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.2 The Effect of an Emergency on Section Capacity . . . . . . . . . . . . . . 54
4.3 The Transport Capacity Calculating Strategy in an Emergency . . . 54
4.4 Calculating Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.4.1 Brief Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
5 Line Planning in Emergencies for Railway Network . . . . . . . . . . . . . 59
5.1 Introduction to Line Planning in Emergencies . . . . . . . . . . . . . . . . 59
5.2 Related Works to Transportation Capacity Calculation . . . . . . . . . . 60
5.3 Modeling for Line Planning in Emergencies. . . . . . . . . . . . . . . . . . 61
5.3.1 Basics of the Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.3.2 Building Line Plan Generating Model . . . . . . . . . . . . . . . . . 63
5.4 Solving the Line Plan Generating Model Under
Condition of Accident . . . . . . . . . . . . . . . . . . . . . . . . . ......... 64
Contents ix

5.4.1 Determining the Frequency of the Different

Types of Trains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.4.2 Designing the Stops of the Trains . . . . . . . . . . . . . . . . . . . . 65
5.5 Simulation Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.5.1 Assumption of the Case . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.5.2 Available Paths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
5.5.3 Frequency Setting and Stops Setting for This Case . . . . . . . 72
5.5.4 Analysis of the Computing Results . . . . . . . . . . . . . . . . . . . 73
5.6 Brief Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
6 Train Re-Pathing in Emergencies Based on Fuzzy Linear
Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6.2 Related Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
6.3 Train Re-Pathing Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
6.3.1 Basic Assumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
6.3.2 Graph-Based Description of Rail Networks . . . . . . . . . . . . . 78
6.3.3 Available Paths Set Generating . . . . . . . . . . . . . . . . . . . . . . 78
6.3.4 Optimization Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
6.3.5 Train Distribution Model. . . . . . . . . . . . . . . . . . . . . . . . . . . 80
6.4 Fuzzy Coefficients Processing and Train Re-Pathing
Model Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.4.1 Fuzzy Coefficients Processing . . . . . . . . . . . . . . . . . . . . . . . 82
6.4.2 Steps to Solve Train Distribution Model . . . . . . . . . . . . . . . 84
6.5 Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.5.1 Case Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.5.2 Trains to Be Re-Pathing . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.5.3 Available Paths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.5.4 Distribution Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.6 Brief Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
7 Train Re-scheduling Based on an Improved Fuzzy Linear
Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
7.1 Introduction on Train Re-scheduling. . . . . . . . . . . . . . . . . . . . . . . . 95
7.2 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
7.2.1 Objective of Train Re-Scheduling Model . . . . . . . . . . . . . . 100
7.2.2 System Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
7.2.3 Mathematical Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
7.3 Fuzzy Linear Programming with Fuzzy Resource Constraints
Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 103
7.3.1 Fuzzy Linear Programming with Fuzzy Resources . . . .... 103
7.3.2 Fuzzy Linear Programming of Resources with Fuzzy
Coefficients Boundaries of Fuzzy Resources . . . . . . . . .... 104
x Contents

7.4 Improved Fuzzy Linear Programming Model

for Train Re-scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
7.5 Computation Cases and Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 107
7.5.1 Results of the Computation Cases . . . . . . . . . . . . . . . . . . . . 107
7.5.2 Analysis of the Computation Cases . . . . . . . . . . . . . . . . . . . 120
7.6 Brief Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Content Summary

This monograph focuses on the train operation theories and methods in emergen-
cies, on the background that the railway accident rate is slightly elevated. It reflects
the authors’ pioneering research work in the field of railway transportation orga-
nization, which is a comprehensive introduction to theories of train operation on the
railway network in emergencies.
This book is structured into seven chapters. Chapter 1 introduces the back-
ground, significance, necessity, contents of the research work and defines the key
concepts involved in this book.
Chapter 2 discusses the train operation in emergencies. It focuses on the analysis
of the connotation and structure of railway transportation organization in emer-
gencies. And it presents the methods and strategies for train operation in
emergencies.
Chapter 3 studies the transportation mode in emergencies.
Chapter 4 discusses the theories and methods for railway carrying capacity
calculation in emergencies.
Chapter 5 discusses the problem of line planning in emergencies for railway
network.
Chapter 6 studies repathing passenger train problem in emergencies.
Chapter 7 researches train re-scheduling problem in emergencies.
This book is rigorously structured and combines the theories and production
reality tightly. It is suitable for the faculty who works or studies in the field of
railway transportation organization. It is also appropriate for graduate students and
researchers in the field of related research.

xi
Chapter 1
Introduction

Abstract This chapter is the introduction on train operation problem during

emergencies. It first defines the train operation problem. Then it introduces some
typical dispatching systems in the world and discusses the possible future railway
operation dispatching mode. Then, we give a demonstration of the significance and
necessity to study this problem. A review on the related publications is proposed in
the second part of it. At last the structure of this book is presented.

1.1 Introduction on Train Operation Problem

in Emergencies

1.1.1 Statement of the Problem

Railway transportation is the most important transportation pattern, while it had

remained undeveloped for a long period. The contradiction between the trans-
portation capacity and requirement is very sharp.
There are mainly two kinds of measures to solve this problem. One is to
strengthen railway infrastructure construction, expand the rail network, and enhance
the transport capacity of the railway system. The second one is to utilize the
advanced technology and management methods and improve the production effi-
ciency of railway transport, which is expected to maximize the transportation
capacity of the existing networks.
At present, the development trends of railway transportation in China are to
expand the high-speed railway construction scale, the gradual separation of pas-
senger and freight transport, and other aspects of automation and intelligent man-
agement. With the high-speed railway mileage increasing, a new railway network
scale is being developed. Since the Tenth Five-Year Plan of the national economy
and social development (2000–2005) was carried out, railway construction has been
accelerated. In January, 2004, the State Council promulgated the “Long-term
Railway Network Plan”, which is a clear blueprint for high-speed railway network
construction. By 2020, China will build 12,500 km high-speed railways and

© Springer Nature Singapore Pte Ltd. 2017 1

L. Jia et al., Train Operation in Emergencies, Advances in High-speed
Rail Technology, DOI 10.1007/978-981-10-4597-4_1
2 1 Introduction

4000 km intercity passenger lines, forming the high-speed railway network, which
will connect all the influential cities—including top 30 largest cities (four munic-
ipalities, 18 provincial capital cities, and eight coastal cities).
In long-term Railway Network Adjustment Plan issued by the central govern-
ment on October 31, 2008, the goal of 2020 national railway operation mileages
was strengthened to 120,000 km, among which the high-speed railway mileage was
changed from 12,000 to 16,000 km. Moreover, the electrification rate is adjusted
from 50 to 60%. In addition, the passenger transportation and freight transportation
are separated on the busy main railway lines. The reasonable layout, clear structure,
perfect function, smooth convergence railway network will be built. Transport
capacity will be able to meet the needs of national economic and social develop-
ment and the quality of main technical equipment will reach the international
advanced level. The modern railway network scale continues to be enlarged, which
provides the hardware basis for the railway transport capacity strengthening.
Chinese railway network is becoming more and more complicated and train
operation organization has become extremely complex.
In addition, at present, the production organization mode of railway is a
three-level management one, which includes railway company, railway bureaus,
and stations levels. Train operation dispatching work is focused on the dispatching
sections. The dispatching mode in view of the whole network is not studied sys-
tematically, which is not harmonious with the complicated train operation dis-
patching work.
On the other hand, China has a vast territory and railway natural disasters are
widespread, frequent, and multiple (Xu 1997). Emergencies have great effect on the
transportation safety and efficiency. The railway accident and railway public
security incidents occasionally occur. These events brought serious challenges to
the railway transportation organization. In emergency conditions, it will cause the
entire railway transportation disorder on the railway line network if the capacity
decreases or even interrupts the situation. Current computer-aided decision system
existing is not able to give a satisfactory solution. However, if the dispatcher gives
the solution subjectively, the quality of the dispatching plan cannot be guaranteed.
So, it is seriously important to study the train dispatching work on the new
railway network in emergencies. Although there is much automatic dispatching
system, none of them can finish the dispatching work on the whole railway net-
work. The goal of this book is to look into the train dispatching problem on the
railway network in emergencies.

1.1.2 Systems on Railway Train Operation Dispatching

At present, the train dispatching systems can be categorized into three groups. The
train dispatching system used in Japan is a typical one of the first group. The train
dispatching system for high-speed railway is separated from the train dispatching
system for normal speed railway. The train dispatching systems in France and
1.1 Introduction on Train Operation Problem in Emergencies 3

Spanish belong to the second group. The novel train dispatching system is derived
from the existing train dispatching system, and combined with the existing train
dispatching system. Train dispatching system in German is the third type, which
builds several local centers for the train dispatching system.
Train dispatching system for Japanese Shinkansen is a separated one. The train
dispatching system is designed and constructed according to the characteristics of
the high-speed railway. It takes the dependence of high risk and operation safety for
the train dispatching system and sings high importance for the safety. And it
considers the high on schedule rate as a most important goal and constructs the
comprehensive dispatching system that integrates many functions.
In Spanish, train dispatching system for high-speed railway is not constructed
perfectly. The idea to develop the high-speed railway dispatching system is con-
strained by the operation rule of the existing train dispatching system. The system
only includes passenger organization, train operation organization, and the dis-
patching work of the locomotives and vehicles. The structure of the system is
simple and the function of the system is not so powerful, which is not compatible
with the high operation speed of the high-speed trains.
High-speed railways are connected with the normal speed railways in German.
A general center and seven subcenters are settled in the railway transportation
system. Passenger transportation and freight transportation are both managed by a
three-level management system, which is dispatching center-sub-dispatching center
and station attendant. All of the centers are connected by the communication system
to exchange the data. In each dispatching center, there are two subcenters, which are
transportation commanding center and the operation control center. The dispatching
system has the function of transportation commanding, automatic conflicts solving,
automatic dispatching, and information service.
In China, train dispatching systems are all separated, for the normal speed
railway, for the intercity railway and for the high-speed railway.

1.1.3 Future Railway Operation Dispatching Mode

The future mode of train dispatching on the railway network has the following
several possible forms.
(1) The railway centralized dispatching mode
The dispatch work of high-speed railway, the normal speed, and the intercity
railway are gathered together in the same department. All of the dispatchers from all
levels will only obey the commands of a relative dispatching console in a single
dispatching system. They do not attend the making decisions work of dispatching
commands, only receive the commands from the dispatching center and transfer
them. This mode is much suitable for the train dispatching work in emergencies for
it is not necessary to transfer the dispatching power when emergencies occur.
4 1 Introduction

(2) The mode that the high-speed railway dispatching system is independent
In this mode, the high-speed railway dispatching system is independent from the
existing normal speed dispatching system. The system should monitor the daily
production site and remotely control the equipment. The special dispatching con-
soles give the dispatching commands to the stations and trains. High-speed trains
must run on the high-speed railway and they are forbidden to transfer to the normal
speed railway. Transportation work of high-speed railway and that of normal speed
railway are completely independent. It is obvious that the merit of this mode is that
the dispatching work is simplified and the dedicated lines are used for the dedicated
trains. Its shortage is also clear that the dispatching power cannot be unified when
emergencies occur, which does not allow trains to share the rails.
(3) Mode that combines the centralized dispatching mode and the high-speed
railway independent mode
In this mode, dispatching systems of high-speed railway, normal speed railway,
and the intercity railway are all independent. In the normal condition, the dis-
patching systems are used for the relative railways, respectively. When emergencies
occur, the dispatching power is unified to a dispatching center, which dispatches all
kinds of trains to operate on the whole railway network, including all kinds of
railways.
We can see that the possibility of using the second mode is relatively small
because of its obvious shortage. The first and the third mode requires the dis-
patchers to do the dispatching work from the view of the railway network, meaning
that we should organize the transportation work for the existing normal speed
railway, high-speed railway and intercity railway. It is necessary to study the train
operation problem based on complex railway network, especially in emergencies.

1.1.4 Significance and Necessity of This Research

Currently research on train re-scheduling problem focuses on the train dispatching

on a single railway section. The researchers have not paid much attention to train
operation research on a railway network, especially in emergencies. It has the
following significance to study the train operation under in emergencies.
(1) The research can enrich the train operation theories on the railway network,
explore the essence of train operation organization scheme in emergencies, and
fulfill the blank of the research field of train operation organization in emer-
gencies. Theories on train operation in emergencies should be a part of train
operation theory system, which can present supporting information for the
railway transportation organizers.
(2) The study can improve the railway transportation efficiency. It is clear that the
transportation organization plan determines the transportation efficiency. Solid
theoretical basis can provide theoretical and method support for transportation
1.1 Introduction on Train Operation Problem in Emergencies 5

plan design. And it can improve the transportation quality, improve transport
efficiency in emergencies, which is expected to solve the problem that emer-
gencies can lead to railway transportation disorder at the condition of lack of
relative theoretical support.
(3) Train operation in emergencies problem must be solved. Natural disasters
affecting railway are characterized by universality, frequency, and variety in
China. Moreover, there are occasional railway accidents railway safety public
incidents, which declined the capacity of the railway line and reduce the safety
and efficiency of the passenger and freight transportation. So, it requires
research on an objective basis to study the train operation problem in
emergencies.
(4) It is infeasible to cancel the passenger trains when the emergency reduces the
railway capacity even break the rail. If the passenger trains (already on the way)
are canceled, the trip of the passengers is suspended, which will bring serious
social influence. Therefore, the dispatchers and the researchers must solve this
kind of problem. It is the requirement of the society.
(5) It can enrich the function of the automatic train dispatching system. Currently,
the automatic train dispatching system works in the normal condition and
provides supporting information for the dispatchers. In emergencies, dis-
patching commands are generated by the dispatchers with their experience. The
results of this research can be embedded into the dispatching system, adding
functions for the train operation in emergencies. It can help to realize the
computer-aided decision-making system of railway operation dispatching
command in emergencies. It is no doubt that the research has a strong practical
significance to the railway transportation dispatching work.

1.2 Related Works Review and Analysis

1.2.1 Related Works on This Problem

This book is concerned about the train repathing and train re-scheduling problems.
So, we list the related works about the two problems.
(1) Research work on train repathing
Research work on train path focuses on support for service plan design. That is
to say, the researchers try to decide the train paths through studying the transfer
facility, travel distance, and travel cost. Cui (1997) studied the train passenger path
generating problem, presenting a method to search for multiple paths. Chen et al.
(2001) studied the optimization of passenger train paths under the premise of a
given railway passenger station layout and train timetable, proposing a 0-1 pro-
gramming model. Wang et al. (2006) hired the genetic algorithm to generate
k-shortest paths based on the analysis of the traditional method to generate the train
6 1 Introduction

shortest paths. Wang and Peng (2007) developed the computer system to realize the
visualization of passenger train paths through the integration of electronic map
technology. Lv et al. (2007) presented a method to generate the train paths with the
constraint of the train number, considering the difference between the path infor-
mation in the TRS and the real passenger path.
The research publications on freight trains path are numerous and the methods in
them can give us much enlightenment. We can categorize the methods in the
publications into mathematical programming methods, heuristic algorithm, neigh-
borhood search method, etc.
A. Mathematical programming methods
Wang and Feng (1996) first defined the car flow paths set and presented the
application method in the car flow management system. Li and Gu (1997) designed
a directed searching method to find the shortest path between every two railway
stations and the computing example proved the efficiency of the algorithm. Sun and
Zhang (1999) defined the overpass network model of railway and gave calculating
steps of DBFS algorithm, which was based on Dijkstra algorithm. Zhou (2002)
designed an algorithm to search for the shortest path from a station to another
station, which was based on the two-fork tree theory. Shi (1996) built a
multi-objective model to solve the car flow path problem, and he and Shi (1999)
carried out two improvements on the GP-STEM algorithm. Shi et al. (1997)
described the relationship between the car flow path and the train formation plan
and proposed a method to optimize comprehensively the car flow path and the train
formation plan. Meng et al. (2008) designed a regularized PCG algorithm for the
multipath assignment model for the development of railway transportation enter-
prises. Jin et al. (2005) analyzed the effect of rapidity, equilibrium degree, maxi-
mization, and other factors on the parallel flow path selection and constructed an
optimization model for train path. Bierlaire and Frejinger (2008) presented a
method to analyze the difference between the real data and the model for train path
selection.
We can see from the current research literature that the mathematical pro-
gramming methods to solve the train path problem are widely used. It is because
that this kind of problem is easy to be described as a linear programming problem
or a nonlinear programming problem with a not very large scale.
When the traffic path problem is combined with the formation plan problem, it
becomes a complex large-scale combinatorial optimization problem, which is
extremely difficult to solve. The researchers often design heuristic method to solve
this kind of problem.
B. Heuristic search method
Heuristic search is a method to evaluate the position of the solution and get the
best solution until the solution reaches the destination. It can avoid much useless
search work and improve the efficiency, gaining the relative excellent solution in a
short period of time.
1.2 Related Works Review and Analysis 7

Lin and Zhu (1996) built the mathematical programming model for the car flow
path optimization and the formation plan problem and designed a simulated
annealing algorithm to solve the model. Lin et al. (1996) also gave a heuristic
algorithm to solve the train pathing problem, considering the manual deciding rules.
Wang et al. (2007) proposed a hybrid genetic algorithm based on stochastic sim-
ulation. Su and Chen (2008) built a 0-1 model and designed a heuristic algorithm
based on damping coefficient concept. Liu et al. (2007) presented a heuristic
algorithm, A* algorithm to solve the train pathing problem. Nong et al. (2010)
designed a Tabu search method to solve the train pathing problem, considering the
problem as a TSP one. Wang and Ji (1999) narrowed the railway network and
presented a heuristic algorithm, taking the distance and transferring satisfactory into
consideration. Zachariadis and Kiranoudis (2010) built a train path model, taking
the minimum cost as the optimization objective and designed a heuristic clustering
algorithm.
Although some heuristic algorithms such as genetic algorithm have higher
computational efficiency, the termination condition is quite difficult to obtain. In
addition, we are not clear whether we have the optimal solution.
C. Other methods
Sun and Jiang (2005) constructed a model to search for the train path, consid-
ering the transportation cost as the optimization goal. The essence of this model is
to find the shortest path with an iteration strategy. Du et al. (2005) built a
multi-objective model to solve the bidirection car flow and empty-loaded car flow
path optimization problem. Xie and Xu (2009) proposed a method to optimize the
railway network and the storage structure of in the computer system. Hong et al.
(2009) designed a two-stage algorithm to solve the train pathing problem. Lee and
Chen (2009) studied the train routing and train timetabling problem, with the goal
to optimize the usage plan of tracks in stations, hiring the local search method.
(2) Research work on train re-scheduling
There are numerous publications on train re-scheduling problem. The train
re-scheduling problem can be divided into two kinds, train re-scheduling on a
single-track railway section and train re-scheduling on a double-track railway
section.
A. Train re-scheduling on a single-track railway section
Train re-scheduling on a single-track railway section was the first problem
drawing attention, and had been a hot research topic for a long period. Szpigel
(1973) built a linear programming model to solve the train-scheduling problem, and
hired the branch and bound algorithm to solve the problem. D’Ariano et al. (2007)
considered re-scheduling problem as a job shop problem. They took the total delay
time as the optimizing goal, constructing a job shop model, solved it with branch
and bound algorithm.
8 1 Introduction

Sauder and Westerman (1987) also built a train re-scheduling model, taking the
total delay time as the optimizing goal and used the exhaustive algorithm to solve
the model. Cao (1994) built a model with optional constraints, and solved it with
dual linear programming algorithm. Zhao (1999) constructed a model that com-
bined the train re-scheduling problem, the usage plans design problem of the station
receiving and sending lines, and the train routing problem. He also utilized the
branch and bound algorithm to solve the model, limited the searching depth and
avoided that the search space is too large. Cheng (1998) proposed a hybrid simu-
lation method, integrating the modern PERT graph theory and time driving method,
to solve the conflicts in train re-scheduling problem. Zhang and Jin (2005) defined
the concept of neighbored trains, taking the deviation function as the optimizing
object. They hired the genetic algorithm to solve the problem. Caprara et al. (2006)
constructed an integer programming model, taking the train operation profits as the
optimizing object, using the Lagrange heuristic algorithm to solve the problem.
Cacchiani et al. (2010) also built a train re-scheduling model on a single-track
section and solved it with Lagrange heuristic algorithm. Li et al. (2008) built the
train re-scheduling model based on the discrete event system theory and solved the
problem with the strategy of Travel First. Liu and Kozan (2009) considered the train
re-scheduling problem as a job shop problem and designed a feasibility satisfaction
method to solve the problem. Zhou and Zhong (2007) solved the constructed train
re-scheduling model with an integrated method of branch–bound algorithm and
heuristic algorithm. It should be pointed out that the researchers began to pay
attention to the uncertainty in train re-scheduling problem. Yang et al. (2009)
provided a goal programming model and solved it with the branch and bound
algorithm.
The key characteristic of train operation on a single-track railway line is that
only one train can occupy a section between two neighbored stations at a same time.
The overpass and cross operation must happen on the stations. So we not only pay
attention to the trains running at the same direction, but also should pay attention to
the trains from different directions.
B. Train re-scheduling on a double-track section
Train re-scheduling problem is also a hot research topic. Araya et al. (1983) built
a 0-1 programming model for train re-scheduling problem on a double-track rail-
way section and designed the branch and bound method to solve the problem. Zha
et al. (2000) used the Lagrange method to solve the train re-scheduling model, with
a linear objective and several nonlinear constraints. Rodriguez (2007) constructed a
constraint programming model, taking the total delay time as the optimizing goal
and hired the branch and bound algorithm to solve the problem. Törnquist and
Persson (2007) proposed a hybrid integer model for train re-scheduling problem,
considering the total delay time and delay cost as the optimization goal.
Iida (1983), Cheng and Qin (1992), Schafer and Pferdmenges (1994) respec-
tively designed the expert system for train re-scheduling problem. Zhang et al.
(2010) constructed a graph theory model and hired the heuristic algorithm to solve
1.2 Related Works Review and Analysis 9

the train re-scheduling problem. The heuristic algorithm is an improved version of

the greedy algorithm. Xia et al. (2008) induced train re-scheduling problems as job
shop dispatching problem. They presented the train path matrix and the operation
order matrix to build a train re-scheduling on a double-track railway section. The
ant colony algorithm determines the drawing order of the train running line and the
timetable is decided by the max-algebra method. Dong et al. (2005) took the
minimal delay rate as the optimizing objective, built a 0-1 programming model to
solve the train re-scheduling problem on a double-track railway section. They also
hired the Tabu search algorithm to solve the model.
Zhang et al. (1998) constructed an experiment system to simulate the train
operation work on the Beijing–Shanghai high-speed railway. Nie et al. (2001) also
built an experiment system to study the train operation on Beijing–Shanghai
high-speed railway and analyzed the strategy for train re-scheduling on high-speed
railway. Li et al. (2006) hired the MAS (Multi-Agent System), described the
re-scheduling problem as a process that multi-subjects occupied the tracks and the
stations. Chen et al. (2002) built a train re-scheduling model, taking the passenger
satisfactory degree as the optimizing objective and used the genetic algorithm to
solve the problem. Wang and Du (2004) also hired the genetic algorithm to solve
the re-scheduling model on a double-track section. Chung et al. (2009) constructed
a hybrid integer programming model, taking the minimum cost of the delay as the
optimizing goal.
Zhao and Dang (2009) improved the genetic algorithm by embedding the chaos
factor and applied it to train re-scheduling problem. Mu and Dong (2010) hired a
three-group particle swarm algorithm to solve the built re-scheduling model on a
double-track section. Jia et al. (2006) graded the re-scheduling problem with the
large-scale system theory and hired the particle swarm to solve the train
re-scheduling model. Cai and Wang (1992) established an expert system for train
re-scheduling problem. Zhou and Zhong (2005a, b) took the total train departure
interval and the total travel time as the optimizing goal and constructed a model for
train re-scheduling problem. They used the branch and bound algorithm. Dorfman
and Medaic (2004) built the re-scheduling model on a double-track railway section,
taking the total operation time as the optimizing goal. Cheng and Yang (2009) also
built a re-scheduling model on a double-track railway section with the total delay
time as the optimizing goal and hired the fuzzy Petri technology to solve the
problem. Jia and Zhang (1993) tried to generate the re-scheduling plan to recover
the train operation work to obey the planned timetable. Zhang et al. (1995) gave a
simulation model of train operation with multi-time clock and multi-station inter-
active program.
Theories and methods for train re-scheduling on a double-track railway section
had many same points with that for train re-scheduling on a single-track railway. In
general, the method also included mathematical programming, soft computing, and
computer simulation. The main difference is that some of the operation rules for
train re-scheduling on a double-track railway section are different with that on a
single-track section, the constraints of the model are different.
10 1 Introduction

C. Train re-scheduling on unclear kind of railway section

Some of the publications did not point out on what kind of railway section,
double-track section, or single-track section, the train re-scheduling problem is
studied. We list the publications here. Jovanovic and Harker (1990) studied the
heuristic technology and hired the hybrid integer programming model to solve the
train re-scheduling problem, which reduced the number of the search points and
improved the efficiency of the algorithm. Chiang et al. (1998) designed a local
dispatching software package to deal with the operation conflicts, which was also a
heuristic method. Li and Zhang (1998) formed the rules database of the expert
system, based on the coordination mode of the inner subjects of the system. Chen
et al. (2009) utilized the rough set theory to solve the train re-scheduling problem.
Qian and Song (2008) also degigned a train re-scheduling model based on the rough
set theory and improved the reduction algorithm for the deciding rules of the rough
set. Petersen and Taylor (1982) used the simulation method to solve the train
re-scheduling problem. Tommii and Satoh (1990) applied artificial intelligence in
the solving work of train re-scheduling problem. Liu (2000) established a train
re-scheduling model based on the Petri network theory and designed a genetic
algorithm for the model. Salim and Cai (1997) solved the overpass station selection
problem with the goal to reduce the cost of the stop plan.
Pu et al. (2001) designed the genetic algorithm to solve the train re-scheduling
problem. Wang et al. (2006) defined the concept of distance between train
timetables and designed the train re-scheduling model based on the penalty func-
tion. He also hired the genetic algorithm to solve the problem. Salido et al. (2007)
designed a constraint satisfactory model with the goal of minimal total operation
time. Abril et al. (2008) constructed a constraint satisfactory model with a super tree
structure for train re-scheduling problem. Ghoseiri et al. (2004) built a
multi-objective model for train re-scheduling problem. The goals are the minimum
total travel time and the minimum energy consumption of the train.
These publications focused on the theories of train re-scheduling. They studied
the design of the objectives and the solving methods of the model. When con-
structing the train re-scheduling model, they did not point out the model was for the
single-track railway or the double-track railway, but described the common con-
straints in a general manner. Even some of the publications did not mention the
constraints of the train re-scheduling problem. These publications usually applied
the newly existed approach at that time in train re-scheduling problem, which were
the supplements of the train re-scheduling theory system.
D. Train re-scheduling on a railway network
There are not many publications on train re-scheduling on a railway network.
Chen and Zhou (2010) took the total travel time as the optimizing goal and built a
train re-scheduling model on a complex railway network. They used the order
optimization theory to solve the problem.
Min et al. (2011) studied the train re-scheduling problem on a double-track
railway network. They took the weighted cost of delay time as the optimizing goal
1.2 Related Works Review and Analysis 11

and presented a column-generation-based algorithm to solve the model. Castillo

et al. (2011) gave a model on a special section, which not only included a
single-track subsection, but also a double-track subsection. Peng et al. (2001)
established an optimization model for the network timetable, presented a method to
decompose the problem. They developed the computer-aided timetabling system.
Ma et al. (2000) proposed a computer-aided system to draw the train timetable
on the railway network. Zhou et al. (1998) designed an algorithm named the
network hierarchy parallel algorithm of train scheduling.This algorithm first pre-
sented the representation of hierarchy nodes of railway network structure and the
representation of sequence events of train timetable,and divided the train sequence
events into groups according to station and section.He presented a new method to
calculate the train timetable which initialized the timetable integrally,graded the
train sequence events by train priority, and triggered the algorithm to shift the
section events’ state according to the state of the station train events on the end
point.
Wu et al. (2008) paper proposed a method of composing a train operation plan
before transport, which can program traffic capacity overall and can make the
transport working in order. By analyzing the requirement of composing a train
operation plan, the paper designed a model of the ARTSR and proposed a pro-
grammed algorithm based on a method of the fixed train path by the regular
characteristic of ARTSR, which could arrange the train schedule on the chosen
transport routes. Kong et al. (2008) studied the military transportation security by
railway in the condition that there occurred a sudden mass transportation require-
ment. The listed the rules in securing the military transportation in emergencies.
Zhang and Yan (2004) built a discrete event model of train vehicle scheduling
problem in emergencies and designed a greedy algorithm to solve the problem.
Wang (2008) provided a time event graph simulation model to optimize train
timetable from the view of a railway network. He (1995) presented a bi-level
programming model for train re-scheduling on a railway network and designed a
hybrid fuzzy particle swarm algorithm to solve the model.

1.2.2 Analysis on the Publications

Publications on train path were mainly about the integration problem of train
pathing and formation plan optimization. They often took the transportation dis-
tance, time, cost, and transportation volume as the optimizing objectives from the
view of transportation operation bureaus. The constraints were the capacity of the
stations and the railway sections. They often built models to optimize the train path
and designed various kinds of algorithms to solve the problem. In addition, pub-
lications on the train pathing problem of the passenger trains often took the travel
time, cost, and satisfactory degree as the goal to provide the optimal path and
transfer plan for the passengers.
12 1 Introduction

Publications usually took the on-schedule rate, total delay time, and the relative
cost as the optimizing objectives when resolving the train re-scheduling problem on
a double-track railway section. The focus is the train re-scheduling model,
including the description of the optimizing goals and constraints and the algorithms.
There are few publications on the comprehensive optimization of train repathing
and train re-scheduling problem, especially in emergencies. Moreover, there are
several publications on train operation on the railway network, focusing the time-
tabling problem and the development of the computer-aided timetabling system. So
the train operation problem should be paid much attention.
There are still some omissions in the present study as follows.
(1) The scope of the study is limited to a single dispatching section
Current publications mostly focus on the study of train re-scheduling problem on
a single railway section, whether it is a single-track railway section or a
double-track section. The essence is to redesign the train operation order and
inbound and outbound time of the trains at stations. Only a few publications are
about the train operation from the view of networked transportation organization.
(2) Lack of study on train operation in emergencies
The present study usually aimed at solving train re-scheduling problem, which
was not necessary to change the train operation paths. The assumed emergency was
not serious and they could not deal with train re-scheduling problem on a railway
network.
(3) The models and the algorithms are rich but not suitable for the problem put
forward in this book.
The approaches can be classified into two types: the optimization approaches
and the simulation approaches. The optimization approaches included mathematical
programming method, heuristic method, soft computing method, expert system
method, etc. The models and the algorithms are quite rich, which give us much
enlightenment. However, this book focuses on train operation on a railway network,
considering the train repathing and train re-scheduling problem comprehensively,
especially in emergencies, which will avoid the inadequacy of studying the prob-
lems respectively.

1.3 Structure of This Book

This book is constructed with seven chapters, See Fig 1.1. Chapter 1 introduces
train operation problem in emergencies and analyzes the related publications in the
research field. Chapter 2 focuses on the theories on train operation in emergencies,
including railway transportation system in emergencies, macroscopic model of
railway transportation, difference between train operation organization and normal
1.3 Structure of This Book 13

Fig. 1.1 Structure of this book

condition in emergencies, and train operation organization strategy in emergencies.

From the third chapter, each chapter studies one subproblem of train operation.
Chapter 3 is about the transportation mode, and Chap. 4 discusses the transport
capacity of the railway section in emergencies. Chapter 5 focuses line-planning
problem in emergencies. Chapter 6 studies train repathing problem in emergencies.
Chapter 7 discusses train re-scheduling problem.

References

Abril, M., Salido, M. A., & Barber, F. (2008). Distributed search in railway scheduling problems.
Engineering Applications of Artificial Intelligence, 21(5), 744–755.
Araya, S., Abe, K., & Fukumori, K. (1983). An optimal re-scheduling for online train traffic
control in distributed situation. In Proceedings of 22nd IEEE Conference Decision and Control
(pp. 489–494). San Antonio, Tex, USA: IEEE.
14 1 Introduction

Bierlaire, M., & Frejinger, E. (2008). Route choice modeling with network-free data.
Transportation Research Part C, 16(2), 187–198.
Cacchiani, V., Caprara, A., & Toth, P. (2010). Scheduling extra freight trains on railway networks.
Transportation Research Part B, 44(2), 215–231.
Cai, B., & Wang, J. (1992). A study on the expert system for train dispatching based on
simulation. Journal of the China Railway Society, 14(3), 31–41. (in Chinese).
Cao, J. (1994). The optimization model and its algorithm for adjusting train diagram on
single-track rail lines. Journal of the China Railway Society, 16(3), 72–78. (in Chinese).
Caprara, A., Monaci, M., Toth, P., & Guida, P. L. (2006). A lagrangian heuristic algorithm for a
real-world train timetabling problem. Discrete Applied Mathematics, 154(5), 738–753.
Castillo, E., Gallego, I., Ureña, J. M., & Coronado, J. M. (2011). Timetabling optimization of a
mixed double-and single-tracked railway network. Applied Mathematical Modelling, 35(2),
859–878.
Chen, D., Li, Y., Zhang, & Gao, S. (2009). Discussion on train re-scheduling approach based on
roughest theory. Chinese Railways, 7, 44–46. (in Chinese).
Chen, Y., Pu, Y., & Jiang, Y. (2002). A study on the satisfactory optimization model and solution
for adjusting train diagram on double-track railway. Bulletin of Science and Technology, 18(6),
463–469. (in Chinese).
Chen, Y., Shi, F., Qin, J., & Zhou, W. (2001). optimization model and algorithm for routing
passenger trains through a railway station. China Railway Science, 31(2), 101–107. (in
Chinese).
Chen, Y., & Zhou, L. (2010). Study on train operation adjustment algorithm based on ordinal
optimization. Journal of the China Railway Society, 32(3), 1–8. (in Chinese).
Cheng, Y. (1998). Hybrid simulation for resolving resource conflicts in train traffic re-scheduling.
Computer in Industry, 35(3), 233–246.
Cheng, Y., & Qin, R. (1992). Research on the expert system for the adjusting of the train diagram.
Journal of the China Railway Society, 14(2), 42–50. (in Chinese).
Cheng, Y. H., & Yang, L. (2009). A Fuzzy Petri Nets approach for railway traffic control in case of
abnormality: Evidence from Taiwan railway. Expert Systems with Applications, 36(4), 8040–
8048.
Chiang, T. W., & Hau, H. Y. (1998). Knowledge-based system for railway scheduling. Data &
Knowledge Engineering, 27(3), 289–312.
Chung, J. W., Oh, S. M., & Choi, I. C. (2009). A hybrid genetic algorithm for train sequencing in
the Korea railway. Omega, 37(3), 555–565.
Cui, B. (1997). Discussion on railway passenger path algorithm. Railway Computer Application, 6
(5), 10–11. (in Chinese).
D’Ariano, A., Pacciarelli, D., & Pranzo, M. A. (2007). A branch and bound algorithm for
scheduling trains in a railway network. European Journal of Operational Research, 183(2),
643–657.
Dong, S., Wang, J., & Yan, H. (2005). Tabu search for train operation adjustment on double-track
line. China Railway Science, 26(4), 114–119. (in Chinese).
Dorfman, M. J., & Medanic, J. (2004). Scheduling trains on a railway network using a discrete
event model of railway traffic. Transportation Research Part B, 38(1), 81–98.
Du, J., Yao, X., & Huang, H. (2005). Satisfactory optimization of the wagon flow problem in
railway network. Systems Engineering, 23(9), 47–49. (in Chinese).
Ghoseiri, K., Szidarovszky, F., & Asgharpour, M. J. (2004). A multi-objective train scheduling
model and solution. Transportation Research Part B, 38(10), 927–952.
He, B. (1995). The transport organization patter for Beijing-Shanghai high-speed railway. China
Railway Science, 16(3), 13–23. (in Chinese).
Hong, S. P., Kimb, K. M., Lee, K., & Parkd, B. H. (2009). A pragmatic algorithm for the train-set
routing: The case of Korea high-speed railway. Omega. International Journal of Management
Science, 37(3), 637–645.
Iida, Y. (1983). Timetable preparation by A.I. approach. In Proceeding of European Simulation
Multiconference (pp. 163–168). Nice France.
References 15

Jia, C., Hu, S., & Yang, Y. (2006). Study on the particle swarm optimization algorithm for train
operation adjustment. Journal of the China Railway Society, 28(3), 6–11. (in Chinese).
Jia, L., & Zhang, X. (1993). Distributed intelligent railway traffic control based on fuzzy decision
making. Fuzzy Sets System, 62(3), 255–265.
Jin, L., Ye, Y., Zhao, Y., & He, J. (2005). Choosing of regional railway network car flow routing.
Chinese Railways, 12, 49–51. (in Chinese).
Jovanovic, D., & Harker, P. T. (1990). Decision support system for train dispatching: an
optimization-based methodology. Transportation Research Record, 1314, 31–40.
Kong, Q., Zhou, F., Wei, G., & Huang, K. (2008). A study of ensuring military railway
transportation in large-scale unexpected demand. Traffic Engineering and Technology for
National Defence, 6(9–10), 14. (in Chinese).
Lee, Y., & Chen, C. Y. (2009). A heuristic for the train pathing and timetabling problem.
Transportation Research Part B, 43(8–9), 837–851.
Li, F., Gao, Z., Li, K., & Yang, L. (2008). Efficient scheduling of railway traffic based on global
information. Transportation Research Part B, 42(10), 1008–1030.
Li, Y., & Gu, S. (1997). Directional searching algorithm for finding shortest path between two
vertexes in railway network. Journal of the China Railway Society, 19(2), 25–27. (in Chinese).
Li, X., Yang, Z., & Du, P. (2006). MAS-Based method for train operation adjustment. Journal of
the China Railway Society, 27(1), 115–119. (in Chinese).
Li, P., & Zhang, Y. (1998). Study on internal coordinating expert system for train traffic regulation.
China Railway Science, 19(3), 1–9. (in Chinese).
Lin, B., Peng, H., & Ren, B. (1996). Optimal weighted car routing in railway network. Journal of
Northern Jiaotong University, 20(6), 645–650. (in Chinese).
Lin, B., & Zhu, S. (1996). Synthetic optimization of train routing and makeup plan in a railway
network. Journal of the China Railway Society, 18(1), 1–7. (in Chinese).
Liu, H. (2000). A study of train control system modeling with petri nets and train operation
adjustment using genetic algorithm. Beijing: China Academy of Railway Sciences. (in
Chinese).
Liu, Z., Ji, L., & Ye, Y. (2007). Study on the shortest path problem in car optimizing in railway
network. Journal of Guizhou Normal University (Natural Sciences), 25(2), 88–90. (in
Chinese).
Liu, S., & Kozan, E. (2009). Scheduling trains as a blocking parallel-machine job shop scheduling
problem. Computers & Operations Research, 36(10), 2840–2852.
Lv, X., Liu, C., Shan, X., & Song, J. (2007). Optimization on route computation algorithm based
on train-route restriction. China Railway Science, 28(3), 122–125. (in Chinese).
Ma, J., Zhou, L., & Hu, S. (2000). Research on train working diagram system of railway netted
lines worked out with computer. Journal of the China Railway Society, 22(1), 7–11. (in
Chinese).
Meng, F., Gao, S., Yang, X., & Wang, Y. (2008). Optimal model for multi-path traffic assignment.
Journal of Liaoning Technical University (Natural Science Edition), 27(supplement), 93–95.
(in Chinese).
Min, Y. H., Park, M. J., Hong, S. P., & Hong, S. H. (2011). An appraisal of a
column-generation-based algorithm for centralized train-conflict resolution on a metropolitan
railway network. Transportation Research Part B, 45(2), 409–429.
Mu, W., & Dong, Y. (2010). Research on train re-scheduling model and Three-particle-swarm
coordination optimization algorithm. Railway Operation Technology, 16(2), 13–15. (in
Chinese).
Nie, L., Zhang, X., Zhao, P., Yang, H., & Hu, A. (2001). Study on the strategy of train operation
adjustment on high-speed railway. Journal of the China Railway Society, 23(4), 1–6. (in
Chinese).
Nong, J., Wang, L., & Yin, H. (2010). Optimization design of railway car flow routing based on
genetic algorithm. Journal of Tongji University (Natural Science), 38(1), 76–80. (in Chinese).
Peng, Q., Zhu, S., & Wang, P. (2001). Study on a general optimization model and its solution for
railway network train diagram. Journal of the China Railway Society, 23(1), 1–8. (in Chinese).
16 1 Introduction

Petersen, E. R., & Taylor, A. J. (1982). A structured model for rail line simulation and
optimization. Transportation Science, 16(2), 192–206.
Pu, Y., Chen, Y., & Pu, S. (2001). A study on the genetic algorithm for model of the train
operation dispatch manage system based on satisfactory optimization. World Sci-tech R & D,
23(6), 56–58. (in Chinese).
Qian, M., & Song, J. (2008). Train operation adjustment based on rough set theory. Journal of
Transportation Systems Engineering and Information Technology, 8(4), 122–126. (in
Chinese).
Rodriguez, J. (2007). A constraint programming model for real-time train scheduling at junctions.
Transportation Research Part B, 41(2), 231–245.
Salido, M. A., Abril, M., Barber, F., Ingolotti, L., Tormos, P., & Lova, A. (2007).
Domain-dependent distributed models for railways scheduling. Knowledge-Based Systems,
20(2), 186–194.
Salim, V., & Cai, X. (1997). A genetic algorithm for railway scheduling with environmental
considerations. Environmental Modelling and Software, 12(4), 301–309.
Sauder, R. L., & Westerman, W. M. (1987). Computer aided train dispatching: Decision support
through optimization. Interfaces, 13(6), 24–37.
Schafer, H., & Pferdmenges, S. (1994). An expert system for real-time train dispatching.
Transactions on the Built Environment, 7, 27–34.
Shi, Q. (1996). Models for rail network system transportation capacity and traffic pathing. Journal
of the China Railway Society, 18(4), 1–8. (in Chinese).
Shi, F., Kong, Q., & Hu, A. (1997). A network method of comprehensive optimization of wagon
path and train formation plan. Journal of the China Railway Society, 19(1), 1–6. (in Chinese).
Shi, Q., & Shi, Y. (1999). Multi-objective linear-programming model and its algorithm for car
flow routing with bidirectional heavy and empty cars in railway network. Journal of the China
Railway Society, 21(1), 1–8. (in Chinese).
Su, S. H., & Chen, Z. Y. (2008). Studies on models and algorithms of optimizing train flow paths
in railway network. Journal Of The China Railway Society, 30(6), 1–6. (In Chinese).
Sun, Y., & Jiang, L. (2005). Route optimizing method of loaded car flow with capacity limit in
railway network. Rail Way Transport and Economy, 27(12), 82–84. (in Chinese).
Sun, W., & Zhang, Y. (1999). Study on both CN model and its algorithm about railway
transportation. Journal of the China Railway Society, 21(5), 106–108. (in Chinese).
Szpigel, B. (1973). Optimal train scheduling on a single line railway. Operational Research, 7,
344–351.
Tommii, N., & Satoh, N. (1990). A train traffic simulation system permitting application of
knowledge engineering. Quarterly Reports of TRRI., 1990, 2.
Törnquist, J., & Persson, J. A. (2007). N-Tracked railway traffic re-scheduling during disturbances.
Transportation Research Part B, 41(3), 342–362.
Wang, Z. (2008). Transportation systems engineering. Nanjing: Southeast University Press.
Wang, Z., & Du, W. (2004). Train operation regulation model for auto-block double line and study
on its heredity algorithm. Rail Way Transport and Economy, 26(6), 61–63. (in Chinese).
Wang, S., & Feng, Y. (1996). Research on region of vehicle flow paths. Journal of the China
Railway Society, 18(supplement), 20–23. (in Chinese).
Wang, B., He, S., Song, R., & Wang, B. (2007). Stochastic dependent-chance programming model
and hybrid genetic algorithm for car flow routing plan. Journal of the China Railway Society,
29(4), 6–11. (in Chinese).
Wang, H., & Ji, L. (1999). Research a rational pathway assemble in our national railway. Journal
of Shanghai Tiedao University (Natural Science Edition), 20(4), 51–54. (in Chinese).
Wang, Z., & Peng, Q. (2007). Optimizing the design of MapX based passenger train operation
routes. Rail Way Transport and Economy, 29(3), 85–88. (in Chinese).
Wang, H., Zhang, Q., Wang, J., Wang, Z., & Zhang, Y. (2006). GA-Based model of train
operation adjustment for high-speed railway. China Railway Science, 27(3), 96–100. (in
Chinese).
References 17

Wu, X., Zhou, L., & Sun, Q. (2008). Study on model and algorithm of emergency train operation
plan. Journal of the China Railway Society, 30(5), 1–7. (in Chinese).
Xia, M., Zhou, L., Sun, Q., & Wang, M. (2008). Study on the train operation adjustment on
double-track line based on ant colony algorithm. Logistics Technology, 27(6), 61–64, 71. (in
Chinese).
Xie, J., & Xu, H. (2009). Research on cars flow path and classification of cars flow algorithm.
Railway Operation Technology, 15(3), 18–21. (in Chinese).
Xu, S. (1997). Cause of formation and types of railway natural disasters in China. Journal of
Suzhou University of Science and Technology, 14(4), 3–39, 47. (in Chinese).
Yan, H. (2004). Study on containers transport organization between railway network container
freight stations. Chengdu: Southwest Jiaotong University. (in Chinese).
Yang, L., Li, K., & Gao, Z. (2009). Train timetable problem on a single-line railway with fuzzy
passenger demand. IEEE Transactions on Fuzzy Systems, 17(3), 617–629.
Zachariadis, E. E., & Kiranoudis, C. T. (2010). An open vehicle routing problem metaheuristic for
examining wide solution neighborhoods. Computers & Operations Research, 37(4), 712–723.
Zha, W., Chen, Z., & Li, X. (2000). Research of theory and method for train operation adjustment
on double-track railway line. Journal of the China Railway Society, 22(1), 12–16. (in Chinese).
Zhang, C., Cao, C., Teng, Y., Liu, S., & Chen, L. (2010). Graph model and heuristic algorithm of
train operation adjustment problem. Science Technology and Engineering, 10(2560–2564),
2573. (in Chinese).
Zhang, X., Hu, A., & Yang, H. (1995). A Train operation simulation model based on constraints of
control in random origin and control in random process. Journal of the China Railway Society,
17(3), 22–27.
Zhang, Y., & Jin, W. (2005). Model and algorithm for train operation adjustment on single-track
railways based on genetic algorithm. Journal of Southwest Jiaotong University, 40(2), 147–
152. (in Chinese).
Zhang, X., Yang, H., Zhu, X., Hu, S., & Hu, A. (1998). Research on simulation experiment system
of Jinghu high-speed railway. Journal of the China Railway Society, 20(4), 1–7. (in Chinese).
Zhao, Q. (1999). An optimal model and its algorithm for adjusting diagram on single-track
railway. Systems Engineering, 17(6), 12–18. (in Chinese).
Zhao, S., & Dang, J. (2009). Study on chaos-improved genetic algorithm for passenger-dedicated
line train operation adjustment. Computer Engineering and Applications, 45(9), 220–222. (in
Chinese).
Zhou, P. (2002). A fast algorithm for finding the shortest path between arbitrary two point in a
traffic road net. Computer Engineering & Science, 19(3), 81–84. (in Chinese).
Zhou, L., Hu, S., Ma, J., & Yue, Y. (1998). Network hierarchy parallel algorithm of automatic
train scheduling. Journal of the China Railway Society, 20(5), 15–21. (in Chinese).
Zhou, X., & Zhong, M. (2005a). Bicriteria train scheduling for high-speed passenger railroad
planning applications. European Journal of Operational Research, 167(3), 752–771.
Zhou, X., & Zhong, M. (2005b). Bicriteria train scheduling for high-speed passenger railroad
planning applications. European Journal of Operational Research, 167(3), 752–771.
Zhou, M., & Zhong, M. (2007). Single-track train timetabling with guaranteed optimality:
Branch-and-bound algorithms with enhanced lower bounds. Transportation Research Part B
Methodological, 41(3), 320–341.
Chapter 2
Theories on Train Operation
in Emergencies

Abstract The destination of this chapter is to introduce the theories on train

operation in emergencies. It first gives a definition of emergency and classifies the
emergencies. Then it analyzes the railway transportation system in emergencies,
including the system composition, boundary, function, evolution, and reconstruc-
tion of the system. It also presents a macroscopic model of railway transport
organization in emergencies, and discusses the difference between train operation
organization and normal condition in emergencies. At last it proposes the train
operation organization strategies in emergencies.

2.1 Introduction on Emergencies

Emergencies and their effects have drawn researchers’ attention from all over the
world. The intension and the extension of emergency are the base to study train
operation in emergencies. Although the definitions of the emergencies are different
today, they all emphasize the expectancy and harmfulness.
A typical definition is given by the European Court of human rights. It says that
emergency is a crisis or a dangerous status, which will affect all of the citizens and
threat the normal life of the whole society (Qi et al. 2006).
Emergency in the United States is called as critical event. It was defined as a
major event under which that in any situation, anywhere in the United States the
need for the federal government intervention to provide supplementary assistance,
to assist state and local governments to save lives and ensure public health, safety
and property or transfer disaster threat (Firenze 2001).
Britain defines emergency as any situation that threatens people’s health, life,
property, and the environment (Yang and Wei 2010).
According to the Emergency response law of the People’s Republic of China,
emergency is a natural disaster, an accident disaster, a public health event, or a
social security incident, which happens suddenly, and cause or may cause serious
social harm, needing to take measures to deal with.

© Springer Nature Singapore Pte Ltd. 2017 19

L. Jia et al., Train Operation in Emergencies, Advances in High-speed
Rail Technology, DOI 10.1007/978-981-10-4597-4_2
20 2 Theories on Train Operation in Emergencies

The generalized contingencies are classified into four categories, as follows:

(1) Natural disasters, including floods and droughts, meteorological disaster,
earthquake disaster, geological disaster, marine disasters, biological disasters,
and forest and grassland fires;
(2) Accidents and disasters, including industrial and commercial enterprises, such
as various types of safety accidents, traffic accidents, public facilities and
equipment accidents, environmental pollution and ecological damage events,
etc.;
(3) Public health events, mainly including the epidemic situation of infectious
diseases, unexplained diseases, food safety and occupational hazards, animal
epidemic situation, and other serious impact on public health and life safety;
(4) Social security incidents, including terrorist attacks, economic security inci-
dents, and so on.
According to the nature of all kinds of public emergencies, severity, controlla-
bility, and the extent of such factors, the overall plan sudden public events is
divided into four levels, namely class I (particularly serious), grade II (Serious),
grade III (critical), and IV (general), followed by red, orange, yellow, and blue said.

2.1.1 Definition and Classification of Railway Emergency

Events

Above is the definition and classification of the general meaning of the emergency.
This book studies the railway train operation organization under the emergency
condition, so the definition of the railway emergency event is given first.
In this book, emergency is defined as an event (the railway accident or railway
public security event) that occurs suddenly on the railway line or station, affecting
the capacity of the railway line or the station, even changing the topology of the
railway network, leading to the result that train cannot operate in accordance with
the planned path and operational timetable. In the event, needing to repeat the trains
and re-schedule the trains to deal with natural disasters to ensure the train arrival
established terminal.
There are three classification methods to classify the railway emergencies
according to different classification criteria.
(1) According to the formation of railway emergencies, the railway events can be
classified into three categories, which are shown as follows.
A. Natural disasters. In this book, natural disaster is a natural event caused by the
astronomy, geography, and other factors, which affect the railway transporta-
tion equipment, or lead to abnormal fluctuations in the railway passenger flow.
It will make the railway could not operate according to the planned timetable
(Wang 2012), including meteorological disasters (such as rain, snow, storm,
etc.), geological disaster, earthquake disaster.
2.1 Introduction on Emergencies 21

Emergency
Natural disaster Reduction or loss of the station Reduction of
events
capacity the railway
Railway accident network
capacity
Reduction or loss of the section
Other emergent event capacity

Fig. 2.1 Relation between the emergency and the effects on the railway network

B. Railway accident. According to Railway Traffic Accident Emergency Rescue,

Investigation and Handling Regulations, the railway accident refers to the
operation accident such as conflict, derailed, fires, explosions, and other effects
of the normal railway traffic accidents in the train operation process. It also
includes railway accidents occurring in the process of operation and the col-
lision accidents with pedestrians, motor vehicles, non-motor vehicles, livestock,
and other obstacles. This book takes this definition as the standard.
From Fig. 2.1, we know unexpected events lead to reduction of railway network
capacity: Natural disasters and accidents on the railway; railway public security
events all can lead to reduction or even loss of the capacity of station and railway
sections. In addition, loss of station capacity causes changes of the railway network
topology.
(2) Classification based on the impact on the train operation (Wang 2012)
A. General railway emergency
General railway emergency event is the one that affects the train operation, but
the effect can be eliminated by re-scheduling methods. It includes the natural dis-
aster, railway accident, and public security event that has little effect on the railway
line, leading to minor fluctuation of passenger flow.
It has the following characteristics.
The duration of the railway line is short, and the capacity loss caused by
emergency can be recovered quickly.
Fluctuation of passenger flow is slight: the growth of the passenger can be
undertaken by the trains in the original plan. According to reference (Wang 2006),
train seat utilization rate is generally taken as 0.7–0.8 to improve the comfortable
degree of the passengers in the daily transportation work. The rate even reached to
be 1.19 in the case of large fluctuations in the holidays. In the case of general
railway emergency, the focus of passengers has shifted to reach to the destination as
soon as possible, rather than the demand for travel comfort. Therefore, the seat
utilization rate can be appropriately increased.
Line capacity loss is sight: disturbs occurred on a section and the trains cannot
run as planned. However, most of the trains can reach their destinations after
re-scheduling work without serious delays.
22 2 Theories on Train Operation in Emergencies

In this case, the original plan can accommodate the growth of the passenger flow
caused by the emergency. It is not necessary to add more trains. Due to the events
of short duration, line capacity loss is small. There is no need to adjust the path and
the running of the train number. We can only adjust the train operation plan on a
time dimension and can finish the transportation job within a reasonable period.
B. Serious incidents
Railway serious incidents refer to the emergent events whose influence on
railway transportation cannot be eliminated through the adjustment of train dia-
gram, including the large loss of local line capacity, larger passenger flow fluctu-
ation and (or) continuous natural disasters which last for a long time, the railway
traffic accidents and (or) public security events. They have the following
characteristics.
The duration is long: the influence lasts for a long period and the capacity loss
caused by an emergency cannot be recovered quickly.
Fluctuation of passenger flow is larger: Although the fluctuation of passenger is
larger, the trains in the plan can accommodate the growth of the passengers.
Line capacity loss is large: Disturbs on the railway cause the rate limiting on most
of the railway sections, or even some of the sections are unavailable which causes the
change of the topology of the railway network. Then the trains cannot run according
to the planned timetable and we cannot complete the transportation task.
In this kind of case, the trains designed in the original plan can accommodate the
growth of the passengers. However, influence of the incidents lasts for a long
period, the capacity loss is large, and we cannot complete the transportation work
only by adjusting the transportation plan on the time dimension. The measure to
reduce the number of trains on the original path should be taken, such as detour,
train reconnection.
C. Pernicious railway incidents
Pernicious railway incidents are the events that are must dealt with the adjust on
the strategy of train operation, and even the cross-industry cooperation, including
the large range of line capacity loss, large passenger flow fluctuations, and (or) the
natural disasters which last for a long period of time, rail transport accidents, and
(or) public safety incidents. The characteristics are as follows.
The duration is long: the influence lasts for a quite long period and the capacity
loss caused by emergency can be recovered after a long period of time.
Fluctuation of passenger flow is larger: the fluctuation of passenger is seriously
large and the trains in the plan cannot accommodate the growth of the passengers.
Line capacity loss is extremely large: disturbs on the railway cause strict rate
limiting on most of the railway sections for a long period of time, or even most of
the sections are unavailable which causes a large change of the topology in the
railway network.
In this kind of case, the contradiction between traffic requirements and transport
capacity is significant. Train re-scheduling and adjustment on the train paths cannot
2.1 Introduction on Emergencies 23

deal with the serious condition. Related companies must cooperate and change the
train operation strategy, and re-establish the train operation plan to release the
contradiction between traffic requirements and transport capacity.
In general, railway bureaus have contingency plans to deal with the natural
disasters and incidents, which cause the decrease of the capacity. However,
emergency plan is only to respond to emergencies, in order to ensure quickly and
effectively carry out emergency rescue operations, reduce accidental loss, and
formulate the plan or scheme. It offers few rules related to the disposal of train
operation organization to support the train operation in emergencies on the railway
network.
Under emergency conditions, train operation organization, which belongs to the
category of operational plan adjustment, is the research works focus of this book.
That is, the work focuses on train operation organization in the condition of net-
work capacity loss.

2.1.2 Characteristics of Railway Emergency Events

According to the analysis above, the characteristics of railway emergency events

are as follows (Wang 2012).
(1) Variety of remote causes
Due to different types of railway emergencies, the formation incentive of the
railway emergency events is also different.
The first category is a natural disaster caused by natural phenomena, such as
heavy rain, blizzard, high wind, earthquake, debris flow, etc., which may cause the
train speed limit or interruption. Second, the status of all kinds of fixed equipment
and mobile equipment in the railway transportation system directly affects the
efficiency and safety of the trains. In addition, some man-made factors, such as the
efficiency of dispatchers and drivers curb personnel and road pedestrian, vehicle
walking path, and other factors, will influence the normal operation of the railway
transportation system.
(2) The sudden nature of the event itself
In recent years, with the development of computer technology, aviation tech-
nology, geology, and other disciplines, people realize the forecast on the temper-
ature, humidity climatic conditions for a period and region. But the most
devastating natural disasters (such as debris flow, landslide, earthquake, etc.) cannot
be forecasted because that the natural disaster has significant uncertainty. In addi-
tion, due to the running state of the equipment, different cycles and the difference of
the technical level, the state of mind and the work efficiency of the railway staff, and
the railway accidents caused by machinery, equipment, and human factors have not
existing rules to follow.
24 2 Theories on Train Operation in Emergencies

(3) Diversity of the objects affected

Depending on the different causes and grades of unexpected events, the affected
people will involve the ones from different industries and fields. First, the influ-
enced scope of emergency events is very different. It may affect the passengers from
one or more lines. In addition, when emergency events cause huge harm and need
rescuing, it will affect the local garrison, police forces, health care, and other
government departments.
(4) Consequences of the emergency events
Railway emergency events will generally cause a greater influence on society. It
may cause a considerable area of the train delay and affect the travel of a great
amount of passengers. Even it may harm the passenger’s life and property, causing
serious damage on railway infrastructure, and greatly reduces the confidence on the
safety of railway transportation.
(5) Urgency of event handling
For delays caused by emergencies, dispatchers should adjust train diagram in
time, reduce the delay spreading, and restore the operation as far as possible.
If the emergency causes serious disruptions, which lead to the break of the rail,
railway operation departments shall immediately organize repair work, adjusting
the train paths. They should seek assistance from the local government if it is
necessary. They can establish emergency rescue agencies at the scene start the
corresponding contingency plans, reduce the effects of the accident. In a word,
regardless of what type and level of emergencies, we should take timely corre-
sponding measures to effectively prevent and control the situation, to reduce the
harm and loss caused by emergency events.
(6) Resonance between the events
Different types of emergencies often influence each other and exist at the same
time in the real operating environment. For example, a wide range of rainfall may
lead to the result that the trains run at a limited speed. At the same time, the rain
may evoke a landslide or debris flow that may cause the interruption of railway
lines. In addition, if is accompanied by thunder and lightning, it can also cause
failure of the railway signal system and bring inconvenience to dispatcher’s work.
The common effect of many kinds of emergencies greatly increases the complexity
of train operation work.
(7) Expansion of the impact scale
The essence of railway transportation is to complete the cargo and passenger
transportation process by occupying the rail lines and station resources. When the
incident causes an abnormal occupancy of a rail line or a station, it is bound to lead
to train delays. Especially under the condition of line break off, it needs to relocate
the trains on the other rail lines to reduce the loss and make full use of network
2.1 Introduction on Emergencies 25

resources. Therefore, a single emergency in a specific section may cause changes of

train operation plan in a network of a region.
(8) Uncertainty of associated attributes
The attributes such as occurring time, influence scope, and duration of an
emergency are difficult to predict and described by precise numbers. The status of
trains also takes on a fuzzy and fuzziness and random show. These characteristics
have increased difficulty of train operation in emergencies.
(9) Difference of traffic organizing content
According to the different extents of the impact of emergency events on the
railway transportation system, the measures to deal with the emergency have much
difference. When the impact of emergency events is relatively small, we can only
adjust train operation plan, which will be able to absorb the interference caused by
emergencies. Thus, we can complete the transportation task within the prescribed
period of time. When the impact of emergency events is large, train operation plan
adjustment time cannot cope with the status. Then we need to adjust the train
operation plan both at the space dimension and the time dimension, namely to
optimize the train number, stops setting strategy, train path and timetable, which
involves two aspects, service plan and train timetable. In addition, in harsh emer-
gency conditions, it often needs to set up an emergency organization and organize
emergency rescue. It is urgent to distribute the passengers as soon as possible. It is
more difficult compared with the first two kinds of circumstances.

2.2 Railway Transport System in Emergencies

2.2.1 System Type of Railway Transportation System

System exists in different forms. Depending on the system-generated reasons and

the properties, systems can be classified into various types. According to the origin
of the system, they can be divided into natural and artificial systems. According to
the size and structure, systems can be divided into simple systems and large-scale
systems. According to the time characteristics, systems can be divided into static
system and dynamic system. According to the relationship between the systems and
the external environment, they can be divided into open system and closed system.
Yan (2004) pointed out that the railway organization scheme of train operation
system is an open artificial system in his doctoral thesis. Naturally, the railway
transport system in emergencies is also an opening of the artificial system. We must
constantly change and optimize the railway transportation system to make the
system adapt to the social environment, which requires railway transportation
system to have sufficient flexibility and adjustability.
26 2 Theories on Train Operation in Emergencies

The railway transportation system is a multi-objective system. The total target of

railway transportation system is to achieve social benefits and economic benefits.
But the specific target is varied and it is very difficult to meet all of the require-
ments, such as efficient, fast, economy, comfort, safety, and environmental pro-
tection, because that there is a very strong trade-off or trade-off phenomenon (Wang
2008). That is, at the same time of a function is optimized, there must be one or
several other functions experience the loss of profits. For instance, there is an
obvious “trade-off phenomenon” between the direct train rate and railway trans-
portation cost.
Such multi-objective conflict phenomenon exists in the railway transportation
system, especially in emergency conditions. So we adjust various goals to obtain
the overall optimal effect of the railway transport system. Especially in the emer-
gency, this kind of adjustment is particularly important.

2.2.2 System Composition

The structure of the system is the interaction of the elements and the way or the
order of each other in the system. It is the specific link between the various elements
of the role of form, which is the internal basis to maintain the integrity of the
system.
Railway transportation system components include equipment and staff. The
equipment is divided into fixed and mobile devices. And fixed equipment includes
railway lines and stations, and mobile devices include trains and EMU. Staff
includes production and management personnel.
Relations between trains and railway line, stations are the most important rela-
tions in railway transportation system. The key work in railway transportation is to
arrange the relation between trains and the railway lines, and the relation between
trains and the stations. The plan is in the form of transportation mode, train service
plan, operation diagram, EMU plan, etc.
Railway transportation system can be divided into staff, stations, rail lines, and
equipment. There are complex relations between the staff, stations, rail lines, and
the equipment of different types of railways. The internal structure of the railway
system is very complicated.
The railway transportation system is a large-scale system. Chinese railway
transportation system includes not only the general speed railway, but also intercity
railway, high-speed railway line, equipment, and staff, which is a system involving
many factors. In emergencies, various types of elements, namely line, equipment,
and personnel, must cooperate and the relationship between these elements
becomes more and more complex. The structure of the railway transportation
system becomes very complicated. Moreover, in emergencies, the railway transport
system elements are with strong random state property changes, which make the
structure of the railway transportation system uncertain. So we can conclude that
the railway transportation system in emergencies is a typical large-scale system.
2.2 Railway Transport System in Emergencies 27

The mathematical description of railway system is Strans ¼ ðSstation ; Rrail ;

Ttrain ; Ffaculty ; RÞ, in which Strans is the railway transportation system, Sstation is the set
of stations, Rrail is the set of railway lines, Ttrain is the set of trains, Ffaculty is the set
of staff, and R is the set of relations between all of the elements.

2.2.3 Boundary of the System

If the railway transportation system is regarded as the analysis object, passenger

demand, the natural environmental conditions constitute the environment of the
system. The boundary of the system is a set of status nodes that can start and end
the action. For the railway transport organization system, system boundary is the set
of OD nodes. That is to say, the nodes are the stations where the passenger flow and
freight flow are generated and disappeared.
Emergency conditions may result in the invalidation of the railway transporta-
tion system in the station, so the boundary of the railway transportation system
under emergency conditions is different from the one under normal conditions.
Under emergency conditions, the boundary of the railway transportation system can
be smaller than the boundary of the railway transportation system under normal
conditions. The mathematical description is as follows.
Under normal conditions, the boundary of the railway transportation system is
Se ¼ Sstation . But under emergency conditions, the boundary of the railway trans-
portation system is SAe ¼ SAstation . And there is SAe Se . SAe is the station set of railway
stations in emergencies.

2.2.4 Function and Behavior of the System

The railway transportation system, whether under normal conditions or under

emergency conditions, plays the role of realizing the displacement of passengers
and cargo. The function of the railway transportation system is determined by the
relationship between the stations, sections, trains, and faculty. Under normal con-
ditions, the function of the railway transport system is visible, which is controlled
by a variety of transportation plans. In emergency conditions, changes occur in the
external environment of the system, which leads to the change of the relationship
between the different parts of the transportation system. The function is affected,
which means that it cannot complete the scheduled transport plan. At this time, it
needs to re-adjust the transportation plan to recover the function of the railway
transportation system. Its mathematical description is as follows:
The railway transportation system, Strans ¼ ðSstation ; Rrail ; Ttrain ; Ffaculty ; RÞ, 8R0 ,
Strans , has the function of F 0 in E 0 environment. The emergency forces the rela-
tionships between different parts of the system to change, then R0 ! R0 . Thus
F 0 ! F 0 . The task is to search for A which can cause the change R0 ! R00 . The
28 2 Theories on Train Operation in Emergencies

function F 0 of system Strans ¼ ðSstation ; Rrail ; Ttrain ; R00 Þ is expected to approach to F 0

as much as possible. That is to say, the emergency changes the function of the
railway transportation system, and then we must find a plan to make the system
remain the function as much as possible through changing the relationship between
the parts of different parts of the railway transportation system.
The work of this book is to find the plan to repath the trains and re-schedule the
trains. Its essence is to design the relationship between train, track, and stations in
railway transportation system. The destination is to remain the function of the
railway system. Train path assignment is the embodiment of the spatial relationship
between train and railway line, but the train re-scheduling plan is the embodiment
of the temporal and spatial relationship between train and railway line.

2.2.5 Evolution and Reconstruction of the System

For any system, the evolution of the system is a basic attribute, and it is an
irresistible trend. The state of the system, structure, behavior, and function change
with the passage of time, which is called the system evolution.
The relationship between the stations, trains, sections, and faculties is the key
relationship in the railway transportation system. With the continuous expansion of
the scale of the railway network, new entrants in the railway system, such as the
stations, equipment, lines, and the staff, form extremely complex temporal rela-
tionships with the existing equipment, lines, and the staff. In fact, the structure is
changed in the railway transportation system.
In addition, the change in the environment can also force the structure of the
railway transportation system change, through altering the relationship between the
elements of the railway transportation system. While the reconstruction is to change
the inner structure of a system as a prerequisite to keep the function of the system,
reconstruction can be divided into two types: the active reconstruction and the
passive reconstruction. Active reconstruction is an active adjustment on the system
structure to stable the system structure and perfect the system function. Passive
reconstruction is an adjustment caused by the emergency effect on the system to
remain the system function.
Under normal conditions, dispatching work on different sections of the railway
line is relatively independent. Staff, equipment, and stations on different lines have
little relationship, except that with a cross-line organization mode, there is a certain
relationship. The coupling degree is weak.
In the event of an emergency, the status of line is affected, forcing the railway
staff to redesign the organization scheme, making the time–space relationship
between trains, lines, and station change. These relations include not only the
relationship between existing trains, railway lines, and stations, but also the rela-
tionship between the high-speed line, the stations, and the existing trains. All these
changes mean the structure of the railway transportation system has been changed.
Its essence is the reconstruction of the railway transportation system.
2.2 Railway Transport System in Emergencies 29

So we can conclude that transportation organization in emergencies is a process

adjusting the inner time–space relationship between the trains and stations
(sections), which is also a process to restrict the system. Detailed embodied form is
the relocation of the trains on the railway lines and the re-scheduled timetable for
the trains.

2.3 Macroscopic Model of Railway Transport

Organization in Emergencies

Generally, we decide to activate the appropriate levels of emergency plans due to

the severity and the extent of emergency. The emergency plan is a macroscopical
plan to deal with the emergency and its effect, which present some constraints for
the design of the detailed operation plan.
So, we first discuss the general principles and methods for the emergency dis-
position, and then establish macro-general suitable model for railway transportation
organization in emergencies.

2.3.1 General Principles and Methods of Emergency

Handling in Emergencies

(1) General principle

First of all, the country establishes a management system for emergency han-
dling of unified leadership, comprehensive coordination, and classification
management.
The State Council under the leadership of the premier will research, decide, and
deploy serious emergencies. According to the actual need, the country establishes
the national emergency command agency, which is responsible for emergency
handling. When it is necessary, the State Council may send a working group to
guide the work.
Local people’s governments shall set up relevant emergency handling depart-
ments, which include the leader of the government, the local Chinese People
Liberation Army, and the Chinese People’s Armed Police Force. The emergency
handling department will lead the emergency handling work and coordinate with
the related departments of Local Government and the departments of a lower level
government to carry out emergency response.
The competent departments of the people’s government at a higher level shall,
within their respective functions and duties, guide, assist the people’s governments
at lower levels and their relevant departments to do a good job in dealing with
emergencies.
30 2 Theories on Train Operation in Emergencies

(2) Emergency handling

After the occurrence of unexpected events, the government should organize all
the related departments and dispatch the rescue teams and social power to deal with
the emergency according to its attributes, characteristics, and serious degree. It
should take emergency measures due to the relative laws, regulations and rules.
A. Disposition of natural disasters or public health events
The government should take the following measures to deal with natural dis-
asters and accident disasters.
a. It should organize the team to rescue and treat the injured persons and evacuate
and properly set the threatened personnel, take other rescue measures;
b. It should control the dangerous sources, indicate the danger area, block the
dangerous places, designate warning area, and implement the traffic control and
other control measures;
c. It should immediately repair the damaged traffic, communications, water supply,
drainage, power supply, gas supply, heating, and other public facilities, provide
shelter and basic necessities of life, and implement the medical and health
measures and other security measures;
d. It should prohibit or restrict the use of equipment, facilities, close or limit the use
of the relevant premises, suspend personnel intensive activities or production
activities which may cause damage, and take other protective measures;
e. It should enable the financial reserves and the reserved emergency relief sup-
plies. And the other relief supplies should be provided when it is necessary;
f. It should secure an adequate supply of necessity for life, such as food, water,
and fuel;
g. It should punish the acts to disrupt the market order, such as store up goods too
much, drive up prices, counterfeit goods, and stabilize the market price, main-
tain the market order.
B. Handling of social security incidents
After the occurrence of social events, the government should organize all the
related departments and dispatch the rescue teams and social power to deal with the
emergency according to its attributes, characteristics, and serious degree. It should
take emergency measures due to the relative laws, regulations, and rules.
a. It should force the use of instruments to fight against each other or the parties
involved in the conflict of violence, and properly resolve disputes in the field to
control the development of the situation;
b. It should control the buildings, vehicles, equipment, facilities, and supply of
fuels, gas, electricity, water, and water in a specific area;
c. It should blockade the relevant premises, roads, and make clear the identity of
the site personnel documents, and restrict the activities of the public places;
d. It should take other necessary measures required by the law and administrative
regulations.
2.3 Macroscopic Model of Railway Transport Organization in Emergencies 31

When the serious social incident happens, the public security organs should
promptly dispatch police power. It should restore normal social order as soon as
possible in accordance with the law and take appropriate coercive measures;
When the emergent event happens and seriously affect the normal operation of
the national economy, the State Council or the relevant competent departments
authorized by the State Council may take safeguard, control, and other necessary
emergency measures to guarantee the people’s basic living needs and minimize the
impact of unexpected events.
The government can requisition emergency rescue equipment, facilities, venues,
vehicles, and other materials from enterprises and individuals when necessary. It
can request other local governments to provide manpower, material resources, and
financial and technical support. It has the authority to require the enterprises to
produce the life necessaries and production necessaries and require public services
departments, such as hospitals to provide the corresponding services.
The government should release the emergency information timely.
The residents’ committee by the emergency site should obey the command from
the superior government and organize the relative people to save themselves or save
each other and stable the social order.

2.3.2 Macroscopic Model of Railway Transport

Organization in Emergencies

The designed macro-model for transportation in emergencies includes three parts of

different levels. The three levels are strategic level, the tactical level, and the
operational level. In this model, the strategic level corresponds the emergency plan,
which will give macro-constraints for the transportation organization model. In this
level, the main work is to decide the train operation mode. In the tactical level, the
model takes the transportation mode as the constraints, which determines passenger
evacuation plan, vehicle usage plan, the train operation section plan, train repathing,
etc. The micro-level is the operational level, which corresponds with train time-
tabling, re-scheduling, EMU usage plan design, etc.
The model is designed in accordance with the principle of top-down order. It has
a distinct gradation, which combines the emergency plan, the handling measure, the
dispatching command, the transportation mode, the service plan, and the timetable.
It is a comprehensive model including all the aspects of transportation in an
emergency. The structure is shown in Fig. 2.2.
Different types and levels of emergencies occur frequently in railway trans-
portation system and trains are affected by varying degrees of interference. Railway
managers and dispatchers take different measures to deal with a different degree of
disturbance. So, managers and dispatchers should follow such treatment flowsheet
in emergencies as shown in Figs. 2.3 and 2.4.
32 2 Theories on Train Operation in Emergencies

First Level
Transportation Mode
Strategic Decisions Emergency Plan
Transportation Pattern Selection
Transportation Pattern Adjustment

Second Level

Service Plan

Passenger Evacuation
Vehicle Usage Plan
Disposition Train Operation Section
Tactical Decisions
Meaures Train Repathing

Third Level
Timetabling and Train Rescheduling

Train Rescheduling
Emergency
Operational Decisions EMU Usage Plan
Dispatching Plan
Crew Planning

Fig. 2.2 Structure of macroscopic model of railway transport organization in emergencies

Fig. 2.3 Treatment flowsheet To judge the

for transportation emergency grade
organization in emergencies

Third grade?

No Yes

Third grade response

Second grade

No Yes

Second grade
response

First grade response

2.3 Macroscopic Model of Railway Transport Organization in Emergencies 33

(1) The third-grade responding process. If the effect of the emergency is slight,
disturbance on trains is also slight and the capacity of the railway line is not
affected seriously. In this occasion, we only need to re-schedule the trains to
complete the transportation task. The problem is a typical train re-scheduling
problem, optimizing the summary total delay time of all the trains, the on
schedule rate, and the passenger satisfactory degree. The affected trains are
from a single railway section. The problem involves only the operational level
in the macro-model for transportation and has no relationship with other levels.
We can optimize the train timetable with the iteration and rolling optimizing
method.
(2) The second-grade responding process. When the third-grade responding pro-
cess cannot deal with the emergency and its effect, the managers will activate
the second responding process. In this occasion, the emergency is more serious
and brings more effect on train operation. Natural disasters or the incidents
reduce the capacity of the railway lines, even lead to rail line interruption. Thus,
not all of the trains can run on its planned path. We must find a new path for the
trains that the path has been interrupted or redesign the running paths for all of
the trains related. In this case, we take remaining the trains in the service plan as
the optimization goal, and consider the cost of the trains, total delay time, the
passenger satisfactory degree, social benefits, etc.
(3) The first-grade responding process. When the effect of emergency is serious
and the normal transportation plan will take a long time to recover, the
second-grade responding is ineffective. The emergency plan should be first
activated to evacuate the passengers. Then we redesign a series of technical
files, including the service plan, timetable, and vehicle usage plan. In this case,
we should design the plans from the top level, strategic level, to the bottom
level, the operational level.

Transportation volume
forecasting
Train paths generating
Transportation volume
Train paths generating
assigning
Train re-pathing Train number calculation
Train re-scheduling or
Train re-scheduling timeabling Timetabling

Fig. 2.4 Train operation organization in different grade emergencies

34 2 Theories on Train Operation in Emergencies

2.4 Difference Between Train Operation Organization

and Normal Condition in Emergencies

The essence of railway transportation is to harmonize different parts of the railway

transportation system, optimize the system inner structure, and improve the trans-
portation system function. Transportation in emergencies is to optimize the dis-
turbed structure by the emergencies and to recover its function in normal
conditions. Today in China, the relationships in the railway system include the
relationship between the normal-speed trains, high-speed trains, and the existing
normal railway lines, the high-speed railway lines. The difference between train
operation organization in emergencies and that in normal conditions is as follows.

2.4.1 Different Goals of Transport Organization

Railway transportation organization includes transportation mode selection, service

plan design, and timetabling. The optimization goals are different at different
organization levels.
(1) Transportation mode level
In normal conditions, when deciding the transport organization mode, we mainly
consider passenger-attracting scope, the safety of transportation service, speed,
punctuality, traffic, and railway operating costs. The transportation organization
mode selection of the pursuit of the goal has obvious diversity.
In the international, France railway system takes the mode that high-speed train
can run on the normal-speed railway line, extending TGV operating mileage. The
optimization goal of transportation mode selection is to expand the accessible range
of high-speed passenger trains and to reduce the number of transfer passengers.
Germany and Japan railway system focuses on the efficiency of passenger
organization when designing the transportation mode, trying to create more direct
transportation conditions and reduce the number of passengers needing to transfer
(He 1995).
In China, as early as 1996, Professor Hu (1996) pointed out that there should be
two kinds of transportation modes on the Beijing–Shanghai high-speed railway.
One is the all high-speed trains run on high-speed railway, and the other is that
high-speed trains and medium trains together run on high-speed trains.
He put forward the transport organization mode of “full speed and rapid transit”,
which is based on the safety, speed, punctuality, and comfort degree of passenger
transport, namely the goal of transportation mode selection is due to safety, speed,
punctuality, and comfort degree.
Peng et al. (2004) considered the utilization of high-speed railway capacity when
they studied the transportation organization mode selection problem. Chang argued
(2008) that we should consider average passenger travel time, average passenger
travel fares, passenger comfort degree, passenger travel convenience degree,
2.4 Difference Between Train Operation Organization and Normal Condition … 35

income of railway bureaus, rail line reconstruction cost, and vehicle costs when
choosing the transportation organization mode.
In emergencies, we should take the transportation task into account. We should
transport the passengers and freights as soon as possible when the emergency
occurs, reducing the effect of the emergencies and the social cost. For example,
Guangzhou Railway Bureau issued an emergency dispatching command in
accordance with the requirements of the Railway General Corporation to deal with
overlay of snow disaster and large passenger flow in the winter of 2008. The
dispatch command required the trains dwelt in Guangzhou station and on the
southern part of the Beijing–Guangzhou Railway to change the path to the Beijing–
Jiulong, Shanghai–Kunming, Sanshui–Maoming, and Jiaozuo–Liuzhou railways.
With the development of the railway network, there may be more chances to change
the train path; then, in the transportation mode, high-speed trains only on
high-speed railway are bound to be changed.
(2) Service plan level
In normal conditions, there are two main optimizing goals when the service plan
is designed. One is the profits of the railway bureaus, that is to say, to control the
operating costs in the condition that the ticket fare cannot float. The other is that the
satisfactory degree, waiting time, transferring time, and total traveling time are
often taken as the optimizing goals from the passenger’s point of view. And some
of the publications combine the goals to construct the optimization model, hiring
bi-level model or multi-objective programming model.
Wang and Yang (2007) divided the operation cost into two kinds, the fixed cost
and the variable cost. They built a model to generate the service plan, which took
the railway cost, seat waste rate, and the total passenger waiting time as the opti-
mizing goals. Shi et al. (2007) presented a bi-level programming model, taking the
railway bureau profits as the upper level model-optimizing goal and the passenger
profits as the lower level model. Chang et al. (2000) proposed a multi-objective
model for service planning of Taiwan high-speed railway. They took the operation
cost and the total traveling time as the optimizing goal.
In emergencies, trains may change the paths to avoid the affected railway sec-
tions. The goal is to reach the destination station as soon as possible. The emer-
gency usually causes the capacity loss and leads to train delay on a large scale. Then
the social service function will be questioned. So the most important objective is to
complete the transportation in emergencies to improve the social benefits. The goal
of the service plan level (tactical level) is to relocate the trains on the affected
railway network and complete the transportation task as soon as possible.
(3) Timetable level
In the normal conditions, the objective is to reduce the total dwelling time, train
delay rate, and total delay time, and to improve the satisfactory degree. Wang et al.
(2007) designed the total dwelling time as the optimizing goal, taking the time-
tabling problem as the cycle event-scheduling problem.
36 2 Theories on Train Operation in Emergencies

In emergencies, the goal of train operation adjustment is different from that of

train operation in normal conditions. In emergencies, the target is to improve the
punctuality rate and reduce the total delay time. In addition, the re-scheduled
timetable is expected to have the recovering capability when it is affected.

2.4.2 Different Orders of Operation Plan

Under normal conditions, train operation organization scheme design follows the
top-down design sequence. The order is transportation organization pattern choice—
service plan design—timetabling—EMU utilization plan design and crew planning.
In emergencies, the division of the hierarchy of the train operation organization is the
same as that of the transportation organization under normal conditions. But the
order of the problem is changed. In emergencies, the train operation organization is
the first to consider whether the operation plan can be adjusted to achieve the
established plan. When adjusting the operation plan fails, we consider changing the
existing operating plan, even transportation mode to complete the transportation
task. Therefore, in emergencies the breakthrough point of the train operation orga-
nization is different from that in normal conditions.

2.5 Train Operation Organization Strategy

in Emergencies

2.5.1 Connotation of Train Operation Organization

in Emergencies

The essence of train re-scheduling in emergencies is to redesign the time–space

relationship between trains and railway lines in the condition that the external
environment is changed. The specific embodiment is the train repathing plan and the
train re-scheduling plan. The problem can be divided into three levels, see Fig. 2.5.
The first level is the coordination problem on the railway network, which is the
highest level. The destination is to relocate the trains on the train paths felicitously
and coordinate the number of trains on the road.
The capacity of the railway is reduced when the emergency affects the railway
line, which cannot undertake the planned transport task. So it is necessary to search
for a number of train paths to share the transport task. First, we search for the paths
and form the feasible path set and provide it for the decision makers. The path set is
the basis to relocate the trains on the available paths. Then we coordinate the
capacity utilization coefficient of all the railway sections due to the train number on
the sections and the capacity. The coordinate result is the relocation plan of the
related trains. Obviously, the prerequisite is that there exist available paths and the
capacities are not zero.
2.5 Train Operation Organization Strategy in Emergencies 37

Fig. 2.5 Hierarchical Train operation organization in

structure of train operation emergencies
organization problem in
emergencies

Network level Train sets on paths

Train group level Train sets in sections

Train level Train

We should relocate the different grades and type of trains on different paths to
gain the most satisfying operation results, considering the equipment and other
resources on the railway lines. The traditional method is to designate temporarily
the paths for the trains, which are usually a subjective decision. The designation
plan is often not very appropriate.
The second level is the one that coordinates the train groups. The trains on
different paths were grouped into different groups according to the grades. Each
group of trains is given a same weight in order to coordinate the operation order of
different groups of trains. That is to say, the priorities of the trains in a same group
are the same, while the trains in different groups have not the same priorities.
The group classification is linked to the grades and the priorities of the trains. In
China, there are various kinds of trains on the railway network. If we divide the
trains according to the transport objects, the trains can be divided into two kinds:
the passenger trains and freight trains. If the trains are divided according to the
operation speed, passenger trains can be divided as high-speed trains, express trains,
fast trains, normal-speed trains, low speed trains, and temporary trains. Freight
trains can be classified into five groups, the parcel express special trains, five
scheduled trains, fast goods trains, coal direct trains, and oil direct trains.
In this book, we classify the trains into seven groups, which have different
priorities in the train dispatching work.
(a) High-speed trains (Started with G)
(b) Intercity trains (Started with C) and EMU (Started with D)
(c) Passenger direct trains (Started with Z)
(d) Express trains (Started with T)
(e) Fast trains (Started with K)
(f) Normal-speed trains
(g) Other trains
38 2 Theories on Train Operation in Emergencies

This book divides the trains to be re-scheduled into seven groups and coordi-
nates the relationship between group and group setting the priority of them.
According to the basic classification of train group, we re-schedule the trains from
the highest level to lowest level due to the group classification, which is called
automatic coordination.
On the railway transport production site, the dispatchers are allowed to reset the
re-scheduling priority of the trains, which is called manual coordination. The reason
is that the priority of the trains may be changed in emergencies. For example, the
relief supplies train must be adjusted as the train with the highest priority.
The third level is the train level. In this level, the priority of each train in a group
is determined, which also includes the automatic coordination and the manual
coordination. Automatic coordination in this level can adjust the priorities of each
train according to the actual situation. For instance, if a train is delayed for a fairly
long period of time, the priority is reduced. Manual coordination may move a train
from one group to another group, or, set the priority of the trains manually.
Currently, research on the train re-scheduling focuses on re-scheduling trains on a
single section and seldom on the train re-scheduling trains on a railway network.
We must consider fully the coordination of the trains on different paths and the
mutual influence relation when studying train re-scheduling on a railway network,
especially in emergencies.
So the connotation of train operation in emergencies is as follows. In emer-
gencies, on the basis of calculating the capacity of the affected railway sections, it
generates the available paths set and re-schedules the trains, including trains relo-
cation plan design, the dwelling time design, and the inbound time and outbound
time at stations of the trains.

2.5.2 Basic View of Train Operation Organization

in Emergencies

Thus, there are two main sub-problems in train operation, train repathing (trains
relocating), and train re-scheduling.
(1) Basic view on train repathing
Due to the general rules of train operation, the higher level trains will be set in
the train timetable; first, the lower level trains and the trains delayed for a long
period of time. In this book, the higher level trains and the important trains that
must be guaranteed are arranged on the better paths, reducing the total delay time as
little as possible. The basic view in train repathing is that the higher level trains
occupy the better paths and relocate the trains on the paths according to the priority.
2.5 Train Operation Organization Strategy in Emergencies 39

(2) Basic view on train re-scheduling

According to the general train operation rule, we guarantee the high-level trains’
punctuality and reduce the total delay time. The newly located trains on the paths
may conflict with the original planned trains on the path. We can discuss the
situation on two cases.
A. The newly relocated trains on the path have the same, or a lower priority,
compared to the originally located trains. Thus, we keep the train running tracks
of the originally located trains to obey the rule of guaranteeing the punctuality
trains. Then the newly relocated trains are inserted into the train diagram to
decide the inbound and outbound time at stations.
B. The newly located trains have a higher priority than the originally planned
trains. In this case, there are two tactics. The first one is to do the work as
presented in the preceding paragraph. The second is to order all of the trains,
including the originally located trains and the newly located trains according to
their level and priority. Then we re-schedule them due to the order. The essence
of this method is to reorder and re-schedule all the related trains.

2.5.3 Basic Method of Train Operation Organization

in Emergencies

The review on the related works tells us that the approaches to solve the train
operation problem can be classified into two types: One is the mathematical pro-
gramming method and the other is the intelligent optimization method. Train paths
generation problem is a mathematical programming problem, which can be solved
with the mathematical optimization method. The train repathing problem is a
complex combinatorial optimization problem, which can be solved through con-
structing a mathematical model. And train scheduling problem is a large-scale
combinatorial optimization problem, whose complexity is much greater than that of
train repathing. Constraints in train re-scheduling work are very complex and many
of them cannot be described by the mathematical formulae. As a result, the problem
cannot be solved with the traditional model and algorithm. The intelligent algorithm
can give the problem a feasible solution in the condition of a certain cost (generally
refers to time and space). Although the approach based on the intelligent algorithm
cannot get the optimal solution, it can gain the satisfactory solution and meet the
requirements of timeliness, which was proved effective. Therefore, we hire
the mathematical optimization method to solve the train repathing problem and the
intelligent optimization method to solve the train re-scheduling problem. So we
combine mathematical optimization method and intelligent optimization method to
solve the train operation problem in emergencies.
40 2 Theories on Train Operation in Emergencies

2.5.4 Principles and Measures of Train Operation

Organization in Emergencies

We should obey the principles and measures when repathing the trains in
emergencies.
(1) Principles and measures for repathing the trains.
A. Higher level trains should be located on the better paths.
B. Higher level trains can transfer from higher level railways to lower level
railways.
C. Lower level trains cannot transfer from lower level railways to higher level
railways, especially high-speed railway.
D. To reduce the transfer between different level railways as much as possible.
E. To meet the stops requirements of high-level trains.
(2) Principles and measures for re-scheduling the trains
A. Higher level trains have the priority to occupy the railway stations and sections.
B. Punctual trains have the priority to occupy the railway stations and sections
when the trains belong to a same group.
C. All the trains depart the stations as early as possible.
D. If a train has a special requirement, it can be arranged with priority.
E. Trains can overtake each other if they belong to a same group.
F. Higher level trains can overtake lower level trains.
G. Lower level trains cannot overtake high-level trains.
(3) Concrete measures for train operation in emergencies
A. To relocate the affected trains on the newly found paths.
B. To accelerate the delayed trains and change the inbound time of the trains at
stations.
C. To speed up the technical operation at the stations and change the outbound
time of the trains.
D. To organize the higher level trains to overtake lower level trains with the
crossover link rails.
E. To organize the trains to run a negative direction.
F. To change the usage plan of the receiving and departing lines.
G. To change the routes of EMUs and extend or shorten the running section of
EMUs.
H. To change the connection time.
Measure A is an essential one for train operation in emergencies, which will
affect the originally located trains on the paths. The implementation process is
complex. Measure B and C are the easiest to realize. Measure D and E are difficult
to implement. Measure D requires the crossover link rail and the related signaling
equipment, or there is the artificial signaling condition. Measure E is the most
2.5 Train Operation Organization Strategy in Emergencies 41

dangerous one because the conflicting routes must be checked and considered.
Measure F has the same complexity and security with Measure E. Measure G and
Measure H have the same complexity and security with Measure A.

References

Chang, H. (2008). Selection of passenger dedicated railway transportation organization mode

based on analytic hierarchy process. Technology & Economy in Areas of Communications, 6
(50), 113–114. (in Chinese).
Chang, Y. H., Yeh, C. H., & Shen, C. C. (2000). A multi-objective model for passenger train
services planning: application to Taiwan’s high-speed rail line. Transportation Research
Part B, 34(2), 91–106.
Firenze, B. (2001). Labor safety system research. Safety Science Journal, 45(2), 31–37.
He, B. (1995). The transport organization patter for Beijing-Shanghai high-speed railway. China
Railway Science, 16(3), 13–23. (in Chinese).
Hu, S. (1996). All-high-speed is the best for Beijing-Shanghai high-speed railway transportation
mode selection. Journal of the China railway Society, 18(A00), 58–80, 90–96. (in Chinese).
Peng, Q., Yan, H., & Wei, D. (2004). On transportation mode of Wuhan-Guangzhou high-speed
railway. Journal of Southwest Jiaotong University, 39(6), 703–707, 711. (in Chinese).
Qi, M., Chi, H., Zhao, H., & Sun, Y. (2006). A research review on public emergency management.
Management Review, 18(4), 35–45. (in Chinese).
Shi, F., Deng, L., & Huo, L. (2007). Bi-Level programming model and algorithm of passenger
train operation plan. China Railway Science, 28(3), 110–115. (in Chinese).
Wang, H. (2006). Study on the train scheme of passenger transport special line. Beijing: China
Academy of Railway Sciences.
Wang, Z. (2008). Transportation systems engineering. Nanjing: Southeast University Press.
Wang, L. (2012). Fuzzy random optimization for train operation in emergency. Beijing: Beijing
Jiaotong University. (in Chinese).
Wang, B., & Yang, H. (2007). The research on the train operation plan of the Beijing-Tianjin
inter-city railway based on periodic train diagrams. Journal of the China Railway Society, 29
(2), 8–13. (in Chinese).
Wang, B., Yang, H., Niu, F., & Wang, B. (2007). Study on model and algorithm of periodic train
diagram generation. Journal of the China railway Society, 29(5), 1–6. (in Chinese).
Yan, H. (2004). Study on containers transport organization between railway network container
freight stations. Chengdu: Southwest Jiaotong University. (in Chinese).
Yang, X., & Wei, H. (2010). Traffic planning management method and its application in
emergencies. Beijing: China Architecture & Building Press. (in Chinese).
Chapter 3
Transport Organization Mode
in Emergencies

Abstract This chapter focuses on the transport organization mode in emergencies.

First, it introduces the various modes of train operation. Then it discusses the
transportation mode selection problem, including the necessity and feasibility of
redesign the transport organization mode. Then, it presents a model of transport
organization mode selection in emergencies.

3.1 Transportation Organization Modes

Transportation organization mode is to determine what kind of trains run on what

kind of railway lines and how to organize the train operation. The transportation
mode on the existing normal-speed railway is that the freight trains and passenger
trains of different grades all run on it. The possible transportation modes of the
newly built railway are all-high-speed mode, high-speed train transferring to
normal-speed railway, passenger trains and freight trains both run on the high-speed
railway and middle speed railway transferring to the high high-speed railway (He
1995).

3.1.1 Different Transportation Organization Modes

of High-Speed Railway of Several Representative
Countries and Regions

High-speed railway is developing fast in recent years and has become the devel-
oping direction in many countries. Because of the different national conditions, the
transportation organization modes vary widely. In France, Japan, Spain, only the
high-speed trains can run on the high-speed railways, while in Germany and Italy,
freight trains and passenger trains are both allowed to run on the high-speed
railway.

© Springer Nature Singapore Pte Ltd. 2017 43

L. Jia et al., Train Operation in Emergencies, Advances in High-speed
Rail Technology, DOI 10.1007/978-981-10-4597-4_3
44 3 Transport Organization Mode in Emergencies

(1) Transportation mode of Shinkansen of Japan

Shinkansen of Japan includes Tokaido Shinkansen, Sanyo Shinkansen, Tohoku
Shinkansen, Joetsu Shinkansen and Hokuriku Shinkansen, which are self-forming
high-speed railway system, respectively. All-high-speed train is the transportation
of Shinkansen. It is also called All-high-speed and Transfer mode. Each train runs
on a high-speed railway and does not transfer from one ling to another.
Non-through passengers must transfer in the system.
Japan railway system kept trying to present better transferring advantage for the
passengers. The railway system used the scribing management after privatization.
Shinkansen and the neighboring existing normal-speed railway are operated by a
same company, which enlarges the attracting range and gains many social and
economic benefits. Paths of the high-speed trains extend to the normal-speed rail-
way and reduce the transfer times of the passengers.
High-speed trains on Shinkansen are the force scattered EMUs. The new main
line has the advantages of high density, large passenger transport capacity, good
security, good service facilities, convenient transfer, convenient passenger travel,
and so on. At the same time, in order to meet the different needs of long-distance
passengers and the passengers along the line, the Shinkansen used a different stops
plan, which has attracted a large number of passenger flows.
(2) Transportation mode of French railway system
French high-speed railway-TGV has three forms of the transportation mode. One
is that the high-speed railway is dedicated for the passenger train. The second mode
is to combine the passenger train and the freight train. The third is to improve the
number of trains and reduce the transfer times. The new passenger dedicated use
refers to allow only TGV high-speed train to run a new high-speed rail, which is the
so-called “pure speed.” Highest speed (the southeast line of 270 km/h) was
improved to 300 km/h (the Atlantic line), and even to 350 km/h (the Mediterranean
line). At the same time, in the case without interrupting operation, they allow the
trains originally running in the passenger dedicated line to transfer to the
normal-speed railway line, and to run at the speed of at least 160 km/h. France
adopted the “high-speed trains transfer to normal-speed line” method, extending the
TGV high-speed train operation distance, expanding its scope of access. It reduced
transfer times of the passengers.
On the whole, the mode of French high-speed rail can be attributed to the full
speed-off-line mode. High-speed rail line is only for high-speed trains. While only
the high-speed trains are allowed to run on the high-speed railway and they are
permitted to extend the path to the normal-speed railway. So they can make full use
of the existing infrastructure and minimize construction difficulties and cost. It is
convenient for the trains to enter big cities, such as Paris and Lyon.
From the overall point of view of the system, France high-speed railway com-
pany took the following organizational measures in order to obtain the best eco-
nomic benefits.
3.1 Transportation Organization Modes 45

A. They set maintenance time according to the operational requirements.

High-speed trains operate during the daytime, and the maintenance time is set
for the night, which is at least 1.5 h.
B. They make full use of the characteristics of TGV high-speed trains that can be
two-way operation. According to the time requirement to turn back they try to
connect a timetable of a direction to another timetable of a counter direction.
C. They make full use of the characteristics of TGV high-speed trains that can be
linked. They join two trains at the peak hour of a day or a week to carry more
passengers.
French railway operation department set the relative train number due to the
passenger flow volume. The mode meets the requirements of the transportation
market and improves the reliability of the railway system, leading to the result that
the attendance rate is high. The number of the stops is reduced and the total travel
time of the trains is controlled. However, the intervals between trains are not
constant. The passengers must prepare a timetable in hand.
(3) Transportation mode of German railway system
German high-speed rail network is formed by the transformation of the existing
railway line (the highest speed is 200 km/h) and the newly built high-speed line
(maximum speed 250–300 km/h). German high-speed railway construction
expresses special emphasis on expanding freight transport capacity, enhancing the
quality of transport, and eliminating the bottleneck of the railway section. As a
result, they use the mode of passenger train and freight train both operating on
high-speed railway. High-speed railway is expected to be very busy.
German ICE EMU uses the periodic transportation mode, in which the intervals
between the trains are constant.
Every 1 h an ICU high-speed train is sent in many big cities of Germany. The
periodic timetable is very easy to remember and convenient for the passengers. For
the railway operation department, the required train number is not large. The regular
operation makes the operating personnel work in order. Thereby, reducing the
irregular operation process. In addition, optimization of maintenance procedures to
reduce the trains to run back. And constant departure interval of train makes it easy
for other transportation vehicle to link up with.
The disadvantage of this mode of transport is the train speed must be compatible
with the train timetable. The result is that the average train speed is reduced. The
passenger trains and freight trains cannot run on the same railway line because of
the short intervals between trains.
To provide more convenience for the passenger transferring, the trains are
designed to stop at a same platform. This mode requires the operation reliability and
punctuality of the whole railway network. In addition, they also apply the method to
exchange the train operation line to reduce the transferring frequency. Although the
high-speed railway system has a good mechanism to make it easy to transfer at
the stations, it still tries to provide the through conditions. For example, some of the
ICE and IC series trains may run at the normal-speed railway.
46 3 Transport Organization Mode in Emergencies

Moreover, German railway system operates the short distance trains to meet the
requirements of the passengers and to improve the flexibility of the service.
An EMU train can operate as a whole on the busy railway line. And it also can be
divided into two parts. Each part can run as a smaller size train, which can be used
on the free railway line. This mode can meet the transportation requirements and
save the railway capacity. It also can compete with the cars transportation.
(4) Transportation mode of Italian and Spanish railway system
IN Italy, freight trains and the passenger trains both operate on the high-speed
railway. The main train type is medium coach. Some of the high-speed trains only
run on the high-speed railway, while other not only run on the high-speed railway,
but also on the normal-speed railway, to reach the big cities that are not along with
the high-speed railway. Additionally, the normal-speed train can run on the
high-speed railway, such as the IC and EC trains, and some of the trains that deliver
fresh and perishable goods also have the right to run on the high-speed railway.
Normal freight trains cannot transfer to the high-speed railway. Madrid-Seville
high-speed railway of Spain is designed to meet the requirements of high-speed
trains and medium speed trains to run together on a same railway line.
(5) Transportation mode of Taiwan railway system
Taiwan railway system takes the All-high-speed and transfer mode. Taiwan
high-speed railway is built with the Japanese Shinkansen as a reference. The highest
speed is 300 km/h, using the power scattered EMUs. The railway service level is
improved greatly through optimizing the transportation mode. Taiwan high-speed
railway to implement leapfrog stop strategy, attracting more passengers along the
high-speed railway line. Starting from Taipei, we can arrive in Kaohsiung by the
high-speed train in 80 min.
(6) Current transportation mode of mainland China system
A. Transportation modes of the railway with the highest speed above 300 km/h
There are six transportation modes for the railway with the highest speed above
300 km/h.
a. Only the self-line trains with the highest speed above 300 km/h can run on the
high-speed railway and there is no transfer operation. The trains needing to cross
the railways all run on the normal-speed railways.
b. Only the self-line trains with the highest speed above 300 km/h can run on the
high-speed railway and there is a transfer operation. The trains needing to cross
the railways all run on the normal-speed railways.
c. The self-line trains with the highest speed above 300 km/h and the trains
needing to cross the railways can run on the high-speed railway. Other trains
needing to cross the railways all run on the normal-speed railways.
d. The self-line trains with the highest speed above 300 km/h and the trains
needing to cross the railways can run on the high-speed railway with the speed
above 300 km/h.
3.1 Transportation Organization Modes 47

e. The self-line trains with the highest speed above 300 km/h and the trains
needing to cross the railways can run on the high-speed railway with a speed
between 200 and 300 km/h.
f. The self-line trains and the crossing trains can run together on the high-speed
railway. Self-line trains must run at the speed above 300 km/h. The crossing
trains can run at the speed above 300 km/h if it is possible and other crossing
trains run at the speed between 200 and 250 km/h.
Among the six modes above, the former four modes are the All-high-speed
modes. Mode e and mode f are the modes that allow different speed trains to run on
high-speed railway. The former four modes are more suitable for the high-speed
railway than the last two in China.
B. Transportation modes of the railway with the highest speed above 300 km/h
Intercity passenger dedicated line is designed to attract intercity passenger flow,
mainly to transport the high density, highly flexible passenger flow, which is a
strong competitor of urban road traffic. The characteristics of its operations are:
Most of the traffic is concentrated in the day. The passengers do not have to wait
when they arrive at the station.
Therefore, the mode of transport organization only allows the self-line intercity
trains. This mode does not consider the link with the normal-speed railway and the
form is quite simple. The trains can be divided into two groups. One is the direct
train, only stopping at the central cities. The other is the train, stopping at each
station. A few trains can transfer on the high-speed railway; both the transferring
condition is very strict. The transportation modes of the trains with the highest
speed of 200 and 250 km/h are as follows.
a. Only the self-line trains are allowed to run on the high-speed line. There is no
transferring operation. The crossover trains all run on the normal-speed railway
line. In this mode, the trains with the highest speed of 200 and 250 km/h are
allowed to run on the high-speed railway. The speed of a direct train is no less
than 250 km/h and the speed of other trains are no less than 200 km/h.
b. At the joint station, the cross train can transfer on the high-speed railway.

3.1.2 Analysis of Different Kinds of Transportation Modes

The modes of railway transportation of the representative countries are summarized

as follows.
A. Only the high-speed trains can run on the high-speed railway and the
high-speed train cannot run on the other kind of railways. This mode is called
All-high-speed and transfer mode.
B. Only high-speed trains are allowed to run on the high-speed railway. And the
high-speed trains can run on the normal-speed railway. This mode is called All-
high-speed and turn to normal-speed railway mode.
48 3 Transport Organization Mode in Emergencies

C. Both passenger trains and freight trains have the right to run on the high-speed
railway, which is called the hybrid transportation mode.
All-high-speed and transfer mode: In this mode, all the high-speed trains must
run on the high-speed railway. Other trains cannot run on the high-speed railway.
The direct passenger number is large and the passengers who need to cross the
railway lines must transfer from a train to another at a certain station. This mode is
suitable for the self-forming passenger dedicated railway lines. The advantage of
this mode is that the train speed is high (the speed can reach to 200–300 km/h), the
interval between trains are short, the operation work is simple and the transportation
capacity is large. However, passengers who need to cross the railways must transfer
one or more times, the travel time is prolonged. So some of the passengers may
choose other transportation modes and increase traffic pressure in the city. So
transfer problem is the key problem of this railway transportation mode.
All-high-speed and turn to normal-speed railway mode: This mode allows the
high-speed train not only run on the high-speed railway, but also on the normal-speed
railway. This mode is suitable for the high-speed railway that is linked with the
normal-speed railway. The advantage is that the speed of the trains on the high-speed
railway are similar. The capacity of the railway is large. The fact that the railway can
run on the normal-speed railway can extend the accessibility of high-speed trains and
expand the service area of high-speed line. It can raise the utilization of high-speed
railway line and reduce the transfer number to solve the problem of passenger
crossing the railways. The shortage of it is that it needs much EMUs to support the
transportation mode and the railways must have the same format.
Hybrid transportation mode: This mode is suitable for the high-speed railway that
was transformed from the normal-speed railway, even from the low speed railway.
The advantage is that the investment is small. But the speed difference between the
trains is large and the railway capacity cannot be fully utilized. The operation work is
more complicated than other transportation mode and the highest speed of the trains
is limited to 160–200 km/h, prolonging the travel time of the passengers.

3.2 Transport Organization Modes in Emergencies

3.2.1 The Necessity of Adjusting Transportation

Organization Mode in Emergencies

When natural disasters and emergent events happen, the capacity of the railway will
be reduced, even the railway line can be broken. The planned transportation mode
may not support the situation, so we need to study how to change or redesign the
transportation mode in emergencies. Otherwise, the train cannot operate as planned
and the railway transportation system will collapse. So it is very necessary to study
the transportation mode redesign problem to distribute the delayed trains and
recover the transportation order.
3.2 Transport Organization Modes in Emergencies 49

3.2.2 The Feasibility of Adjusting the Transportation

Organization Mode in Emergencies

(1) Matching feasibility of Train and railway line

The high-speed railway with the design highest speed of 200–250 km/h uses the
ballasted tracks. Then the normal-speed trains, the normal-speed locomotives can
transfer to this kind of high-speed railway. They can run at the speed of their limited
speeds.
The high-speed trains with the speed of 200 km/h can run on the high-speed
railway with the design highest speed of 300 km/h or the railway with the slab
tracks. Other level of trains can run on the high-speed railway if they change the
freight bogie to assure they are not harmful to the rail tracks. In the same manner,
the high-speed trains can run on the normal-speed railway if the car wheels are not
damaged.
(2) Feasibility of Train operation control
The rules of the CTC system are as follows:
A. The train over speed protection system conforming to the CTCS specification
should be able to meet the requirements of the whole process control of a set of
vehicle equipment.
B. System onboard equipment is downward compatible.
C. System level conversion can be completed automatically.
D. System ground, vehicle configuration, such as the condition, in the system fault
conditions should be allowed to downgrade the use.
E. System level conversion does not affect the normal operation of the train.
F. System at all levels should have a clear representation.
Specific requirements are:
A. CTCS level conversion happens in the railway section (which should not
happen at the home signal), and provides the corresponding sound and light
warning signals to the driver. The driver presses the confirming button, lift the
warning signal. When automatic conversion fails, the driver completes
the conversion manually according to the respective warning information about
the ATP onboard equipment and LKJ.
B. CTCS level conversion should be set up with a fixed information responder that
has the function of forecast and implement. Each operation direction needs to
be set separately the responder. The implementation responder can be used with
the interval fixed responder.
C. In the conversion between the levels, the control of the car should be ensured to
get a reliable and smooth transition. The power to control the train is based on
ATP vehicle equipment.
50 3 Transport Organization Mode in Emergencies

D. If it has been the starting point for braking, the braking action should be
maintained. Then the driver can complete the automatic conversion.
To sum up, we can realize the conversion between CTCs at all levels of the
system to a certain extent when the trains running in the railway sections, to ensure
that the trains can run on different railways in emergencies.

3.2.3 The Model of Transportation Organization Mode

Selection in Emergencies

Railway transport organization mode is designed to determine the space relation-

ship between different types of train, between the train and the network. Elements
of the system are mainly railway network and EMUs, then the relationship between
the network and EMUs is the logical relationship between the internal elements of
the system relationship.
In emergencies, the target of the transportation organization is to quickly
evacuate the passengers. The objective function of the model of the railway
transport organization problem is to maximize the passenger evacuation speed.

TOP ¼ ftop1 ; top2 ; . . .; topn g ð3:1Þ

(1) Objective of transportation mode selection model

A. Macroscopic transportation organization mode
The transportation mode can be described as

TOP ¼ HD  NU ð3:2Þ

where

HD ¼ fhd1 ; hd2 g ð3:3Þ

hd1 represents that the high-speed train can transfer to normal-speed railway;
hd2 represents that the high-speed train cannot transfer to normal-speed railway.

NU ¼ fnu1 ; nu2 g ð3:4Þ

nu1 represents that the normal-speed train can transfer to high-speed railway;
nu2 represents that the normal-speed train cannot transfer to high-speed railway.
3.2 Transport Organization Modes in Emergencies 51

B. Medium transportation organization mode

a. If divide the classified the high-speed trains and the normal-speed trains, then

Y
M Y
N
TOP ¼ HTi  NT j ð3:5Þ
i¼1 j¼1

The number of the transportation mode is 2M þ N , n ¼ 2M þ N where M is the

number of the high-speed train types, N is the number of the normal-speed train
types.

HTi ¼ fhti1 ; hti2 g ð3:6Þ

NTi ¼ fnti1 ; nti2 g ð3:7Þ

hti1 represents that the ith type high-speed train can transfer to normal-speed
railway;
hti2 represents that the ith type high-speed train cannot transfer to normal-speed
railway;
nti1 represents that the ith normal-speed train can transfer to high-speed railway;
nti2 represents that the ith normal-speed train cannot transfer to high-speed
railway.

C. Another approach is not to distinguish the high-speed trains and normal-speed

trains. The trains will be divided into M types, and then, we use the relationship
between the train and the lines to express a variety of transport organization
modes.

[
TOP ¼ TRAIN  L; ð3:7Þ

where

TRAIN ¼ ftrain1 ; train2 ; . . .; trainM g ð3:8Þ

L ¼ fl1 ; l2 ; l3 g ð3:9Þ

l1 represents that trains can only run on the high-speed railway;

l2 represents that trains can only run on the normal-speed railway;
l3 represents that trains can run both on the normal-speed railway and the
high-speed railway.
Then, the number of the transportation modes is 3M, n ¼ 3M.
The objective of transportation selection model is
52 3 Transport Organization Mode in Emergencies

top ¼ optimalftop1 ; top2 ; . . .; topn g ð3:10Þ

That is
X X
minTwaiting ¼ noriginal
i ti þ ntransfer
i titransfer ð3:11Þ

(2) Constraints of the transportation mode selection model

A. The existing railway structure. Whether there are links between the high-speed
railway and the normal-speed railway

N intersect ¼ HN \ NN 6¼ £; ð3:12Þ

where HV ¼ fhv1 ; hv2 ; . . .; hvN g represents the set stations nodes on the
high-speed railway,NV ¼ fnv1 ; nv2 ; . . .; nvN g represents the set stations nodes on
the normal-speed railway.

HVdisabled  N intersect ð3:13Þ

B. The constraints from the disabled stations and sections

Disabled stations nodes cannot play the role of a crossing station node. The set
of the disabled station nodes is the proper subset of the joint stations of high-speed
railway and the normal-speed railway.

Reference

He, B. (1995). The transport organization patter for Beijing-Shanghai high-speed railway. China
Railway Science, 16(3), 13–23. (in Chinese).
Chapter 4
Calculation of Railway Transport
Capacity in an Emergency Based
on Markov Process

Abstract This chapter proposes a method to calculate the railway transport

capacity in an emergency to support the dispatching work of control centers. The
effect of an emergency on section transport capacity is analyzed and the basic
method to calculate the transport capacity is presented. The results show that the
situation-changing process in emergencies is actually a Markov process. The cal-
culation rule is presented based on division of the time segment during which the
emergency lasts. The algorithm is designed to calculate the transport capacity of
each time segment. The pessimistic strategy and the fuzzy strategy are proposed to
determine the computing value of the transport capacity of each time segment, to
satisfy the calculating requirements in different occasions. Our study shows that the
method is reasonable and practical and the method can be embedded in the train
dispatching system to support the operation work in an emergency.

4.1 Introduction

The section capability of a railway is the maximal number of trains that can go
through the section in a period of time (a day or an hour), with certain kinds of
locomotives and cars, and with a certain kind of organization rule. From the def-
inition, the section capability is related to the equipment, such as locomotives and
cars and the organization plan. In reality, the locomotives and cars are relatively
stable. So the section capability is mainly determined by the organization rule and
plan. The organization rules cover a series of organization constraints and leading
principles, such as the minimum running time through the sections, the minimum
intervals between trains and the maximal speed that the section can support. The
most significant factor that affects the transport capacity is the maximal speed in the
section.
There are three major methods to calculate the section capacity. An analytical
method is designed to model the railway environment by means of mathematical
formulae or algebraic expressions (Zhao 2001; Hu 1991; Abril et al. 2008).
A simulation method simulates the trains’ dynamic behavior of a system by moving

© Springer Nature Singapore Pte Ltd. 2017 53

L. Jia et al., Train Operation in Emergencies, Advances in High-speed
Rail Technology, DOI 10.1007/978-981-10-4597-4_4
54 4 Calculation of Railway Transport Capacity in an Emergency …

it from state to state in accordance with well-defined rules to get the capacity (Yang
and Yang 2002; Chang 2009). An optimization method is more strategic for solving
the railway transport capacity problem and provides much better solutions than
purely analytical formulae (Steven 2009; Oliveira and Smith 2000). In this chapter,
an analytical method is proposed, which is different from the methods in reference
1–3 in that it uses the Markov process to describe the dynamic changes of segment
capacity.

4.2 The Effect of an Emergency on Section Capacity

The transportation organization pattern, the ratio between different types of trains
and similar information is all uncertain. While we cannot calculate the actual
capacity, we can calculate the transport capacity of the parallel train graph.
An emergency compels the control center to adjust the speed constraint according
to the degree of the emergency, which leads to the reduction of the section capacity.
The most popular formula to calculate the transport capacity is:

1440  60  ðTs þ Tw Þ
N¼ ; ð4:1Þ
I

where N is the capacity, Ts is the skylight time which is reserved to do the main-
tenance work (s), Tw is the extra wasted time (s) and I is the interval between two
neighbor trains. So the calculation of I as shown below is the key point.

Lt þ Ls þ Lp þ Lb
I¼  3:6 þ Ta ð4:2Þ
v

where Lt is the length of the train (m), Ls is the length of time segment that former
train occupies (m), Lp is the length to assure the following train no to enter the
section the preceding train occupies (m), Lb is the braking distance (m), v is the
average speed of all the trains (km/h), and Ta is the additional time reserved for
drivers to recognize the signals. Different braking distances correspond with dif-
ferent speed constraints.

4.3 The Transport Capacity Calculating Strategy

in an Emergency

The degree of an emergency changes randomly during a period of time.

Accordingly, the maximal speed will be changed by the control center. So, the
section transport capacity also changes with the change of the degree of the
emergency. We prove that the changing process is a typical Markov process.
4.3 The Transport Capacity Calculating Strategy in an Emergency 55

We divide the time section into N time segments, as shown in Fig. 4.1. The maximal
speeds in each time segment are possibly different, which leads to the different capacity.
Set
fDti ¼ ti  ti1 ; i ¼ 1; 2; . . .; N g ð4:3Þ

Then
X
N
cT ¼ ai ; ð4:4Þ
i¼1

where ai is the actual section transport capacity of the time segment i.

Although the lengths of every time segment are not the same because the length
is determined by the actual situation of the emergency, we assume that they are the
same in this chapter. We will study the relation between the time segment length
and the emergency in the future.

4.4 Calculating Algorithm

Suppose that there are M constraint speeds in an emergency in a railway. Relatively,

there are M transport capacity values (marked as c1 ; c2 ; . . .; cM and
c1 \c2 \    \cM ) that form the transport capacity value set C.
The constraint speed situation-changing matrix is
0 1
p1;1 p1;2  p1;M
B p2;1 p2;2  p2;M C
B C
P¼B . .. .. C: ð4:5Þ
@ .. . . A
pM;1 pM;2 pM;M

c
a1 aN-1

a2 aN

...

0 t1 t2 t N-1 tN

Fig. 4.1 Changing process of section in an emergency

56 4 Calculation of Railway Transport Capacity in an Emergency …

The period of time that the emergency lasts, T, is divided into N time segments.
Each time segment corresponds with a constraint speed. During T, the emergency
situation changes for N − 1 times. The transport capacities of each time segments
are a1 ; a2 ; . . .; aN . Their values are all from set C. a1 is the transport capacity of the
first time segment. Set a1 ¼ ci , 1  i  M. Then
X
M
Eða2 Þ ¼ pi;1 c1 þ pi;2 c2 þ    þ pi;M cM ¼ cj pi;j ð4:6Þ
j¼1

Set of former time segment of kth time segment as ch ; h ¼ 1; 2; . . .; M. That is to

say, ak1 ¼ ch , then the equation can be generalized as follows:
X
M
Eðak Þ ¼ ph;1 c1 þ ph;2 c2 þ    þ ph;M cM ¼ ch ph;j ; l ¼ 2; 3; . . .; N: ð4:7Þ
j¼1

The transport capacity values of N − 1 time segments except the first one can be
calculated with Eq. 4.7, while h is not the same. That is to say, the expected
capacity of i + 1 time segment is determined by its former situation, the ith situ-
ation. It is not related to the former time segments (the i − 1 time segments)
situation of ith situation. We can see that the process of the situation changing is
Markov process.
The value of transport capacity must be a determined value, not an expected
value. The control center sets the speed constraint at the beginning of every time
segment. So the transport capacity of the time segment is defined. And value of
each time segment can be found in set C. The capacity of each time segment can be
calculated with the equation as follows.
There must be two elements in set C, which meet the inequality ci  Eðak Þ  ci þ 1
where Eðak Þ is the expected value of ak . Then the actual value of ak must be either ci
or ci þ 1 . There are also two strategies to determine the value of ak .

Strategy 1: Pessimistic strategy

Set ak ¼ ci . That is to say, the pessimistic strategy is to set the transport capacity
value to be the smaller one, which probably forces the control center to find more
paths for trains, assuring the reliability of the train operating plan.

Strategy 2: Fuzzy strategy

For that capacity has the fuzzy characters in an emergence, the capacity can be
calculated based on fuzzy functions. The fuzzy functions tending to ci and ci þ 1 are
designed, respectively. A possible fuzzy membership function tending to ak is as
follows:
8
<0 Eðak Þ ¼ ci þ 1
ci þ 1 Eðak Þ
r1 ¼ ci \Eðak Þ\ci þ 1 : ð4:8Þ
: ci þ 1 ci
1 Eðak Þ ¼ ci
4.4 Calculating Algorithm 57

We get

ci þ 1  Eðak Þ
r1 ¼ : ð4:9Þ
ci þ 1  ci

By the same rule, the membership function tending to ci is

Eðak Þ  ci
r2 ¼ ð4:10Þ
ci þ 1  ci

Then

ci r1  r2
ak ¼ ð4:11Þ
ci þ 1 r1 \r2

It can be seen that the triangle fuzzy membership function is used to calculate the
section transport capacity in the emergency. The essence of this method is a
closer-choosing rule. Of course, we can design other functions to replace the tri-
angle function, such as trapezoid function, Gaussian function, Bell function and
Sigmoid function.

4.4.1 Brief Summary

We analyze the randomness of the emergency when designing the method. The
changing process of an emergency situation is proved to be a Markov process. The
pessimistic strategy to determine the actual transport capacity of a time segment can
assure the reliability of the operating plan. And the fuzzy strategy can make the
calculated results as close to the actual value as possible and the algorithm is easy to
realize. The method can be embedded in a train dispatching system to support the
operation work in an emergency. We will study the time segment division and the
dynamic capacity pattern and probability in the future.

References

Abril, M., Barber, F., Ingolotti, L., Salido, M. A., Tormos, P., & Lova, A. (2008). An assessment
of railway capacity. Transportation Research Part E, 44(5), 774–806.
Chang, H. (2009). A simulation study of installation locations and capacity of regenerative
absorption inverters in DC 1500V electric railways system. Simulation Modeling Practice and
Theory, 17(5), 829–838.
Hu, S. (1991). Railway Section Capacity Calculation of Former Federal Republic of Gemany, 3,
18–20. (in Chinese).
58 4 Calculation of Railway Transport Capacity in an Emergency …

Oliveira, E., & Smith, M. B. (2000). A job-shop scheduling model for the single-track railway
scheduling problem, Research Report 2000.21, Leeds, England: University of Leeds.
Steven, H. (2009). Capacity factors of a mixed speed railway network. Transportation Research
Part E, 45(5), 830–841.
Yang, Z., & Yang, Y. (2002). A study on parameters for calculation block section carrying
capacity and trains coefficient of removal on Beijing-Shanghai high-speed railway. Journal of
Northern Jiaotong University, 23(5), 11–17. (in Chinese).
Zhao, L. (2001). Calculation and analysis of carrying capacity of high-speed railway section.
China Railway Science, 22(6), 54–58. (in Chinese).
Chapter 5
Line Planning in Emergencies
for Railway Network

Abstract To generate line plan in emergencies for railway networks to complete the
passenger transportation, we first build a mathematical model in this chapter,
focusing on the frequency setting and stops setting. Then, considering the OD
passenger flow data, we first propose the method to solve the train frequency setting
problem of different types. Genetic algorithm is designed to solve the stops setting
problem. The approach is tested with the data from the Beijing–Shanghai high-speed
railway and its neighbor existing railway. We find that the model is suitable to
generate line plan in emergencies for railway networks and the algorithm has good
calculating performance. The new method to generate line plan proposed in this
chapter can be embedded in the decision support system for railway operators.

5.1 Introduction to Line Planning in Emergencies

Train planning plays a critical role in operating and managing railroad systems. The
planning problem faced by every railway operator consists of several consecutive
stages, ranging from strategic decisions concerning, e.g., the acquisition of rolling-
stock, to operational traffic control. Strategic problems are largely driven by esti-
mates for the long-term demand and constrained by the capacity of the lines. Line
planning is the tactical step of the whole planning process as shown in Fig. 5.1,
which follows the basic step––demand estimation and capacity calculation. And
line planning problem can be divided into three steps, train pathing, train frequency
setting, and train stops setting.
High-speed railway is developing very fast today, which has already improved
the topology structure of the railway network. Thus, a modern railway network is
being built, including the newly built high-speed railway and previously built
railway. Accordingly, there are two types of rails, high-speed rail and normal-speed
rail, and there are links between them at some important railway intersections,
providing the possibility to allow the trains to transfer from one kind of rail to
another. In addition, it provides more space to permit the railway managers to route
the trains more freely.

© Springer Nature Singapore Pte Ltd. 2017 59

L. Jia et al., Train Operation in Emergencies, Advances in High-speed
Rail Technology, DOI 10.1007/978-981-10-4597-4_5
60 5 Line Planning in Emergencies for Railway Network

Demand Line Rolling stock

Timetabling Shunting
estimating planning planning

Capacity Crew
calculating planning

Fig. 5.1 Steps of the whole planning process of railway transportation

Moreover, natural disasters affect railways with characteristics of universality,

frequency, and variety in recent years. Furthermore, railway accidents occur
occasionally, which reduce the capacity of the railway line and degrade the safety
and efficiency of the passenger and freight transportation. Objective research into
the line planning problem in these emergencies is therefore required.
In this chapter, we focus on the line planning problem on the railway network. In
Sect. 5.1 we take a review of the research publications on the problem. In Sect. 5.2,
we build the model to generate the line plans under condition of disasters. We
propose the method to solve the model built in Sect. 5.3. A computing case is
presented to analyze the model correctness and the effectiveness of the algorithms
in Sect. 5.4. In the last section, we draw the conclusion.

5.2 Related Works to Transportation Capacity

Calculation

A line refers to a cyclic schedule of a set of trains on a particular route and direction
of travel. City traffic (metro) lines are typically planned for a day duration and
long-distance passenger traffic lines are planned for anywhere between a day to a
week. Line planning is considered a strategic railway operation because of the
longer planning phase. There is much research work on the line planning problem.
Guihaire and Hao (2008) presented a global review of the crucial strategic and
tactical steps of transit planning: the design and scheduling of the network. They
separated the papers on line planning into four groups according to different
approaches.
The first group uses the mathematical approaches. Murray (2003) studied two
variations of line planning problem. In the first part, the relocation of stops stations
in an existing network is considered with the objective of minimizing the number of
stops. The second part deals with the optimal location of stops to create or extend
the network. Given a fixed number of additional stops to locate, the objective is to
maximize the extra service access provided to non-covered areas. An integer linear
programming model is developed for line planning to satisfy customer demands.
A relaxation approach using branch and bound heuristic is also proposed (Bussieck
1998). Lindner (2000) proposed a cost optimal model for line planning using a
branch and bound method. Carey and Lockwood (1995) presented a strategic model
and algorithm using branch and bound for pathing an additional train in an existing
schedule.
5.2 Related Works to Transportation Capacity Calculation 61

The second group exploits the heuristic approaches. Patz (1925) was probably
the first to tackle the transit network design problem using heuristics. He put
forward an iterative procedure to generate a lines network using penalties. Initially,
the network contained a line for each origin–destination pair. Sonntag (1977)
presented a heuristic procedure originally created for the line planning problem of
railway systems. Mandl (1979) tackled the line plan starting with an empty routes
network. He proposed a heuristic algorithm to define a transit network given a
constant frequency on all bus lines. Jovanovic and Harker (1991) reported heuristic
and metaheuristic approaches to scheduling of railway traffic. Ghoseiri and
Morshedsolouk (2006) also proposed a heuristic approach to solve the train
scheduling problem, utilizing colony system. Michaelis and Schöbel (2009) inte-
grated line planning, timetabling and vehicle scheduling and designed a customer-
oriented heuristic approach.
Neighborhood search approaches are introduced in the third group of publica-
tions. An aggregated metaheuristic approach to the transit network design problem
is considered by Zhao and Ubaka (2004), Zhao and Zeng (2006) with the objective
of minimizing the number of transfers and optimizing route directness while
maximizing service coverage. The concept of keynote is defined to elaborate
neighborhoods in the context of met heuristics solution methods. An integrated
simulated annealing, Tabu and Greedy search algorithm is proposed by Zhao and
Zeng (2006) while basic greedy search and fast hill climb search are implemented
by Zhao and Ubaka (2004). These algorithms were tested on benchmark instances
and data from Miami Dade County, Florida.
The fourth group uses evolutionary algorithms. Xiong and Schneider (1993)
presented an innovative method to select supplementary routes for an existing
network. Their method is based on an improvement on the ordinary genetic algo-
rithm, called the cumulative GA (Genetic Algorithm). Chakroborty and Dwivedi
(2002) also proposed a genetic algorithm based on this method. An initial set of
routes are determined heuristically, then a process which consists of an evaluation
and modification procedure is repeated to obtain the optimal solution.
References provide us many approaches to solving the line planning problem
under conditions of disasters and give us much important inspiration on building
the model and designing the algorithm.

5.3 Modeling for Line Planning in Emergencies

Much research work has been done on the transportation organization mode of the
high-speed train. There are two patterns under the normal condition.
(1) The high-speed trains run only on the high-speed railway lines. They cannot
transfer from high-speed rail to the normal speed rail and the normal-speed
trains cannot transfer from normal-speed rail to high-speed rail.
62 5 Line Planning in Emergencies for Railway Network

(2) The high-speed train can switch from high-speed rail to normal-speed rail and
run jointly with the normal-speed train, but the normal-speed trains cannot
transfer from normal-speed rail to high-speed rail.
Emergencies may break the high-speed railway and affect the train operation
seriously. Therefore, it is necessary to allow the trains to switch to normal-speed
railway under such situation. In addition, the links between the high-speed railway
and the normal-speed railway become essential.

5.3.1 Basics of the Model

The sets and variables are listed below.

W the number of the stations in the railway network
M the number of the segments in the railway network
S node (station) set
E edge (segment) set
O a subset that covers all the stations that can play the role of starting station
D a subset that covers all the stations that can play the role of terminal
station
qu;v OD flow between station u and station v
B the number of the train grades
k the grade of the trains determined by running speed
qu;v
k the number of passengers transported from station u to station v by k-type
of train
lu;v
k the number of the k-type trains from station u to station v
Ack the seating capacity of the k-type train
ri the capacity of station i
ni the capacity of segment i
ku;v
k;i a flag variable to denote whether k-type of trains pass through station i
gu;v
k;i a flag variable to denote whether k-type of trains pass through segment i
dku;v The distance of the path that k-type trains run from station u to station v
Tturnover time consumed to finish the valid passenger transpiration
u;v
tk;r time consumed for k-type trains running from station u to station v
u;v
tk;d time consumed for k-type trains dwelling at stops on the path from station
u to station v
1u;v
k the number of the k-type available rolling stock from station u to station v
The railway network is described with G ¼ ðS; EÞ. k denotes the grade of the
trains separated by the speed.
We take it for granted that the same kind of trains take the same path from a
starting station to another. Then we set
5.3 Modeling for Line Planning in Emergencies 63

1; kth type trains pass through station i;
ku;v
k;i ¼ ð5:1Þ
0; kth type trains do not pass through station i:


1; kth type trains pass through station i;
gu;v
k;i ¼ ð5:2Þ
0; kth type trains do not pass through station i:

5.3.2 Building Line Plan Generating Model

(1) Optimizing objective

There are three main objectives for the line plan generating model. One is the
profit of the railway bureau. The other is the profit of the passengers. The detailed
objectives are as follows:
A. To maximize the profits of the railway bureau. The profit is related to the tickets
revenue, the transport device loss, and the cost in the train operations.
B. To minimize the cost of the travel for the passengers. The cost of the passengers
includes time and money, which make it a requirement for the passengers that
they spend the minimum money and minimum time to complete the travel for
the line plan. The problem concerns the ticket price, stops plan, transferring
time, line connection, and comfort degree on trains.
C. To maximize the number of passengers served. It is related to the coefficient of
utilization of the railway capacity.
We take the passenger distribution efficiency as the optimizing objective in the
emergencies. When an accident happens, it will decrease the transport capacity and
may create gigantic passenger backups. Meanwhile, the safety of the passengers
should be guaranteed. Thus, profits of the railway bureau are no longer the opti-
mizing objective. To distribute the passengers as quickly as possible is the goal.
Passenger distributing efficiency is defined in Eq. (5.3).
PB
0 qu;v
k dk
u;v
Q ðu; vÞ ¼ k¼1
ð5:3Þ
T turnover

Tturnover denotes time consumed to finish the valid passenger transportation,

which can be calculated in Eq. (5.4).
XXX B XXX B
Tturnover ¼ Trun þ Tstop ¼ lu;v t
k k;r
u;v
þ lu;v u;v
k tk;d ð5:4Þ
u2O v2D k¼1 u2O v2D k¼1
64 5 Line Planning in Emergencies for Railway Network

where Trun is the time consumed by all the trains to run from their own starting
station to terminal station; Tstop denotes the time consumed to make a stop at the
stations;
(2) Constraints
There are many constraints in the line planning problem, especially in emer-
gencies. We choose the most important constraints carefully as follows:
X u;v
qk ¼ qu;v ; u 2 O; v 2 D ð5:5Þ
k¼1;2

qu;v
k
 1:3Ack ; u 2 O; v 2 D; k ¼ 1; 2 ð5:6Þ
lu;v
k

XXX
B
ku;v u;v
k;i lk  r ;
i
i ¼ 1; 2; . . .; W ð5:7Þ
u2O v2D k¼1

XXX
B
gu;v u;v
k;i lk  n ;
i
i ¼ 1; 2; . . .; M ð5:8Þ
u2O v2D k¼1

lu;v u;v
k  1k ; u 2 O; v 2 D; k ¼ 1; 2 ð5:9Þ
þ
lu;v
k 2 N ; u 2 O; v 2 D; k ¼ 1; 2 ð5:10Þ

Equation (5.5) denotes that the number of passengers allocated to each train is
equal to the number of passengers forecasted to be transported. Equation (5.6)
ensures that the load ratio of the trains cannot exceed 130%. Equation (5.7) denotes
that the station capacity of approaching trains must be bigger than the designed
sending task in the line plan. Equation (5.8) requires that the number of trains
passing through a certain segment designed in the line plan must be smaller than its
technical capacity. Equation (5.9) is a transportation resource constraint that
requires the train number cannot be bigger than the number of rolling stock.
Equation (5.10) requires the number of the trains to be a positive integer. N þ is the
set of positive integers.

5.4 Solving the Line Plan Generating Model Under

Condition of Accident

We divide the solving process into two steps. The first one is to calculate the train
numbers of different types. The second is to determine the stops of the trains.
5.4 Solving the Line Plan Generating Model Under Condition of Accident 65

5.4.1 Determining the Frequency of the Different Types

of Trains

(1) Design of the types of trains

The railway network is constructed by the high-speed railway and normal-speed
railway, and the trains are divided into two types, the high-speed trains and the
normal-speed trains.
The high-speed trains can transfer from the high-speed railway to normal-speed
railway and the normal-speed trains are not allowed to transfer on the high-speed
rail and the intercity rail, taking the capacity coefficient of utilization into consid-
eration. Moreover, the speed of high-speed trains may be affected by the rails that
they run on.
(2) Calculating the numbers of trains of different types
There is a widespread rule to design the number of the trains on the segments.
That is to calculate the number of the trains according to the number of the pas-
sengers. The equation to calculate the number of the trains is as follows:

qu;v
lu;v
k 
k
ð5:11Þ
ð1 þ aÞAck

In emergencies, a certain degree of overload is allowed on the train to transport

the passengers as quickly as possible. a is the overload ratio. We set the overload
ratio to be 0.3 according to the Chinese railway passenger transportation policy.
Then the equation is innovated to be

qu;v
lu;v
k 
k
ð5:12Þ
1:3Ack

Generally, the train number must be an integer, so the k type train number is
 
qu;v
lu;v
k ¼ INT
k
þ1 ð5:13Þ
1:3Ack

where INT is a bracket function.

5.4.2 Designing the Stops of the Trains

It is very hard to select the stops of a train since the stops selection is influenced by
numerous factors. We use a genetic algorithm to solve it, with 0 representing that a
train does not stop at a station and 1 denoting that the train stops at a station. The
66 5 Line Planning in Emergencies for Railway Network

searching space is very large and there are many decision variables. The heuristic
algorithm is, therefore, suitable for solving this problem. The steps of the algorithm
are as follows.
(1) Design the code of the problem;
(2) Initialize the population Xð0Þ ¼ ðx1 ; x2 ; . . .; xL Þ;
(3) Calculate the value of adaptive function FðxL Þ of each chromosome in popu-
lation XðtÞ;
(4) Create the middle generation XðtÞ;
(5) Create the new generation Xðt þ 1Þ based on the middle generation XðtÞ;
(6) Set t ¼ t þ 1; If the exiting condition does not exist, go to step (4).
The algorithm designed for the train stops setting problem is as follows:
(1) Coding of the problem
Coding is to express the feasible solution as a characteristic string, which can
describe the characteristics of the problem. And the codes are required to be easy to
deal with. The train stops can be coded as a string and there is a 0 or 1 at each bit.
0 denotes that the train will not stop at a certain station and 1 denotes that it will.
When the strings of all the trains are linked, a chromosome is formed.
In this chapter, the code of k type trains stops is designed as a one-dimensional
array xLk . The length Lk is

L ¼ mu;v
k ð5:14Þ

mu;v
k denotes the number of the stops when k-type trains run from station u to
station v.
Then the population size of the chromosome is

XXX
B
pop size ¼ lu;v
k mk
u;v
ð5:15Þ
u2O v2D k¼1

The code of a chromosome is constructed as shown in Fig. 5.2.

(2) Initialization of the population
Initialization of the population is to construct the original population as the initial
solution to the problem. To create the population of the initial solution is to generate
pop_size chromosome, which is shown in Fig. 5.2. And it is a constraint to meet the
station capacity requirement. So it is necessary to judge whether the capacity
constraint is satisfied during the process of generating the initial solution.

1 0 ... 0 1 1 0 ... 0 1 ... 1 0 ... 0 1 ... 1 0 ... 0 1 1 0 ... 0 1 ... 1 0 ... 0 1 ...
1,2 1,2 1,3 1,3 1,2 1,2 u ,v u ,v
l1 m1 l1 m1
u ,v
l1 m1
u ,v
l2 m2 1,3
l2 m2
1,3
l2 m2

Fig. 5.2 Coding representation

5.4 Solving the Line Plan Generating Model Under Condition of Accident 67

(3) Evaluation of the population

It is essential to evaluate the population to judge the quality of the population.
The index is the value of the adaptive function. The adaptive function value is the
symbol of the adaptability of the chromosome. The bigger the function value is, the
more opportunities to survive the chromosome will have. The adaptive function and
the objective function are closely related to each other. The objective function can
be seen as the mutation of the adaptive function. Generally speaking, the adaptive
function can be used as the objective function when the objective function is
nonnegative or the problem is to maximize the objective function value. When the
goal is to minimize the objective function value, the objective function can be
designed as the difference between a large number and the adaptive function. In this
chapter, passenger distributing speed is taken as the objective function and it is
nonnegative. It is a maximal optimizing problem, so the adaptive function is
designed to be equal to the objective function.
(4) Selection
Selection is to select an outstanding chromosome from a generation and pass its
excellent gene to the next generation. The roulette method is often adopted in
programming. The roulette method is a proportion strategy. Its main idea is to select
the chromosome according to the adaptive function value. The four steps of roulette
are as follows:
A. Calculate the adaptive value evalðch Þ of every chromosome ch ;

evalðch Þ ¼ f ðxÞ; h ¼ 1; 2; . . .; pop size ð5:16Þ

B. Calculate the summary of the adaptive value of all the chromosomes;

X
F¼ evalðch Þ ð5:17Þ
1  h  pop size

C. Calculate the selecting probability ph of chromosome ch ;

evalðch Þ
ph ¼ ; k ¼ 1; 2; . . .; pop size ð5:18Þ
F

D. Calculate the accumulated probability rk of every chromosome ch .

X
rh ¼ pj ; h ¼ 1; 2; . . .; pop size ð5:19Þ
1jh

Then a plate is formed and it is cycled for pop size times. Every time one
chromosome is selected to create the new generation.
68 5 Line Planning in Emergencies for Railway Network

A. Generate a pseudorandom number z 2 ½0; 1;

B. If z  r1 , then select the first chromosome c1 , otherwise, to select the hth
chromosome ch to meet the requirement rh1  z  rh .
(5) Crossover
The crossover operation includes one-point crossover, multi-point crossover,
part- crossover, etc. In this chapter the one-point crossover is hired. The operation
steps are as follows:
A. Select a single point as the starting point randomly from two father
chromosomes;
B. Exchange values at the same places from the starting point till the end of the
chromosomes, generating two different chromosomes.
(6) Mutation. The task of the mutating is to change the value of a specified bit. In
this chapter, mutating is to negate on the bit. That is to say, to change 1 into 0
and 0 to 1. It has two steps. The first step is to select a place to determine the bit
to execute mutation operation. The second is to execute the negating operation.
Then the new generation is created after step (2) to step (6).
(7) Judge if the loop can be canceled. If yes, then exit, or else go step (2).

5.5 Simulation Example

5.5.1 Assumption of the Case

We suppose that a serious accident happens at the segment on the high-speed rail
between Xuzhou and Bengbu. The task is to design a line plan for the next day. The
passenger OD flows between the stations in this accident are calculated and fore-
casted based on the historical data and the negative effects of the accident at that
time. The OD flows data are listed in Table 5.1. The numbers of rolling stock
available for the paths in the accident are shown in Table 5.2. The emergency will
last for 3 days and it requires the running speed to be reduced to 40 km/h in section
between Xuzhou and Bengbu. Other sections are not subject to the emergency. The
window time is set to be 240 min in the emergency for the workers to overhaul the
railway equipment to assure safety. Under normal condition, the interval time
between the departure times of two trains must be longer than 6 min on the normal
rail in Chinese railway system. The maximum speed is 160 km/h. And the interval
time on the high-speed rail must be longer than 5 min. The maximum speed is
350 km/h. According to the method that we proposed (Meng et al. 2012), we
calculated the sections capacities and got the stations capacities with the help of the
dispatchers from Shanghai Railway Bureau. The capacities are shown in Table 5.2.
In this case, number of train grades B = 2.
 5.5 Simulation Example

Table 5.1 Forecasted OD flows of passenger in a day between stations in the accident
Xuzhou Shangqiu Bengbu Xinyi Yangzhou Fuyang Huainan Hefei Nanjing
Xuzhou – 9644 13,918 4548 246 1205 247 2164 85,808
Shangqiu – 4329 2164 494 767 2630 2877 4548
Bengbu – 115 0 98 493 3836 6339
Xinyi – 493 88 0 130 493
Yangzhou – 79 90 110 1918
Fuyang – 5507 4795 1671
Huainan – 4740 1644
Hefei – 4329
Nanjing –
69
70 5 Line Planning in Emergencies for Railway Network

Table 5.2 Basic data (a) Number of rolling stocks available for the paths in the
of the railway network accident
in the emergency
Running segments Number
of rolling
stocks
H N
Xuzhou–Shangqiu–Fuyang–Huainan–Hefei– 0 10
Nanjing
Xuzhou–Xinyi–Yangzhou–Nanjing 0 10
Xuzhou–Bengbu–Nanjing 60 5
Shangqiu–Xuzhou–Bengbu–Nanjing 0 5
Shangqiu–Xuzhou–Bengbu 0 10
Bengbu–Huainan–Hefei 0 10
Fuyang–Huainan–Hefei 0 10
Hefei–Nanjing (high-speed rail) 15 5
Xuzhou–Shangqiu 0 10
Xuzhou–Bengbu (high-speed rail) 0 10
Xuzhou–Xinyi 0 12
Xuzhou–Shangqiu–Fuyang–Huainan–Hefei 0 8
Shangqiu–Xuzhou–Xinyi 0 5
(b) Section capacity in the accident
Railway segments Capacity
Xuzhou–Xinyi 98
Xuzhou–Shangqiu 200
Xuzhou–Bengbu (normal-speed rail) 200
Xuzhou–Bengbu (high-speed rail) 50
Xinyi–Yangzhou 98
Yanzhou–Nanjing 98
Shangqiu–Fuyang 98
Fuyang–Huainan 98
Huainan–Hefei 98
Heifei–Nanjing 200
Bengbu–Huainan 98
Bengbu–Nanjing (normal-speed rail) 200
Bengbu–Nanjing (high-speed rail) 240
(c) Station capacity in the accident
Railway segments Capacity
Xuzhou (on normal-speed railway) 125
Xuzhou (on high-speed railway) 298
Xinyi 65
Yangzhou 72
Nanjing (on normal-speed railway) 180
Nanjing (on high-speed railway) 320
(continued)
5.5 Simulation Example 71

Table 5.2 (continued) Shangqiu 80

Fuyang 165
Huainan 72
Hefei (on normal-speed railway) 125
Heifei (on high-speed railway) 240
Bengbu (on normal-speed railway) 106
Bengbu (on high-speed railway) 160

5.5.2 Available Paths

The railway lines around are shown in Fig. 5.3. Paths found for the trains are as
follows:
• Xuzhou (high-speed rail)–Bengbu (normal-speed rail)–Nanjing
• Xuzhou (normal-speed rail)–Bengbu (normal-speed rail)–Nanjing
• Xuzhou (high-speed rail)–Bengbu–Huainan–Hefei (high-speed rail)–Nanjing
• Xuzhou (high-speed rail)–Bengbu–Huainan–Hefei (normal-speed rail)–Nanjing
• Xuzhou (normal-speed rail)–Bengbu–Huainan–Hefei (high-speed rail)–Nanjing
• Xuzhou (high-speed rail)–Bengbu–Huainan–Hefei (high-speed rail)–Nanjing
• Xuzhou–Shangqiu–Fuyang–Huainan–Hefei (high-speed rail)–Nanjing
• Xuzhou–Shangqiu–Fuyang–Huainan–Hefei (normal-speed rail)–Nanjing
• Xuzhou–Xinyi–Yangzhou–Nanjing
According to the paths above and the stations characteristics, we design the OD
pairs that can be the starting stations and terminal stations, which are as follows:
• Xuzhou–Nanjing
• Xuzhou–Shangqiu
• Xuzhou–Bengbu
• Xuzhou–Xinyi
• Xuzhou–Hefei
• Shangqiu–Nanjing
• Shangqiu–Bengbu

Fig. 5.3 Affected part of the

railway network. Note The
thicker lines stand for
high-speed railway lines and
the thinner lines stand for
normal-speed railway lines.
The numbers besides the
lines are the lengths of
the railway sections
(measurement unit: km)
72 5 Line Planning in Emergencies for Railway Network

• Bengbu–Hefei
• Fuyang–Hefei
• Hefei–Nanjing
• Shangqiu–Xinyi
The paths between the OD pairs are shown at the left side of Table 5.3.

5.5.3 Frequency Setting and Stops Setting for This Case

In this chapter, a high-speed train contains 16 cars and its seating capacity is 1200.
The normal-speed train contains 17 cars and has the seating capacity of 1400. We
first get the forecasted OD flows of passenger data between stations, shown in
Table 5.1. According to the methods of the previous section, we calculate the
number of the two types of the trains based on the passenger data. The computing
results are shown in Table 5.3. We set the crossover probability to be 0.6 and
mutation probability to be 0.1. The maximal iteration number is 1000.
The adaptive function raised value to its peak value 263,100 after 580 iterations.
Then we get the stops plan through interpreting the best chromosome. The stop plan
is shown at the right side of Table 5.3.

Table 5.3 Train operation plan for segment Xuzhou–Bengbu

Running segments H N Stop plan
Xuzhou–Shangqiu–Fuyang–Huainan– 0 3 One NST stops at Fuyang, one NST
Hefei–Nanjing stops at Hefei, the last has no stops
Xuzhou–Xinyi–Yangzhou–Nanjing 0 3 One NST stops at Yangzhou, one NST
stops at Xinyi. The other has no stops
Xuzhou–Bengbu–Nanjing 50 4 8 HSTs stop at Bengbu
Shangqiu–Xuzhou–Bengbu–Nanjing 0 3 All stop at Xuzhou
Shangqiu–Xuzhou–Bengbu 0 3 All have no stops
Bengbu–Huainan–Hefei 0 2 One stops at Huainan
Fuyang–Huainan–Hefei 0 6 3 NSTs stop at Huainan
Hefei–Nanjing (high-speed rail) 9 0 All have no stops
Xuzhou–Shangqiu 0 3 All have no stops
Xuzhou–Bengbu (high-speed rail) 0 5 All have no stops
Xuzhou–Xinyi 0 4 All have no stops
Xuzhou–Shangqiu–Fuyang–Huainan–Hefei 0 3 2 NSTs stop at Shangqiu
Shangqiu–Xuzhou–Xinyi 0 2 Both have no stops
Note H represents high-speed trains and N represents normal-speed trains in this table. NST
represents normal-speed train. HST represents high-speed train
5.5 Simulation Example 73

5.5.4 Analysis of the Computing Results

According to the computing results, the high-speed trains are located for the OD
pairs of Xuzhou and Nanjing, Hefei and Nanjing. This is because that high-speed
rail exists between the two pairs of OD stations. And we can see that there is no
normal-speed train from Hefei to Nanjing because there are not so many passengers
OD from Hefei to Nanjing and the goal is to improve the transportation speed.
Other OD pairs all have the normal-speed trains to meet the passengers’ journey
requirements of going out.
50 high-speed trains will be sent from Xuzhou to Nanjing. The foremost reason
is that the OD flow between Xuzhou and Nanjing is great that requires so many
trains to transport the passengers. Another reason is that the goal is to improve the
transportation efficiency, so the number of high-speed trains is 50, which is much
bigger than the number of the normal-speed trains. Furthermore, we can see that the
number is equal to the capacity of the section between Xuzhou and Bengbu. The
solution makes extensive use of the section capacity.
On the path Xuzhou–Shangqiu–Fuyang–Huainan–Hefei–Nanjing, 3 trains are
designed to finish the passenger transportation task. Two of them are set to stop at
the intermediate stations. On the path Fuyang–Huainan–Hefei, there are six trains
allocated, and three of them have a stop at Huainan. This is because that the OD
flow between Fuyang and Huainan is 5507 and the OD flow between Huainan and
Hefei is 4740. These are quite large OD flows. For the same reason, two of three
trains designed on the path Xuzhou–Shangqiu–Fuyang–Huainan–Hefei stop at
Shangqiu station.
From the analysis above, we can see that the computing results are reasonable
which can prove the validity of the model. It demonstrates that we can generate the
feasible line plan for a local railway network in emergencies with the model and the
algorithm.

5.6 Brief Summary

In this chapter, a method is presented for line planning in emergencies for the
railway network. We divided line planning into two steps. The first is to calculate
the number of different types of trains. The second is to determine the stops of the
trains along the railway line.
The above computational results and analysis show that it is practical to use the
model presented to describe the line planning problem on the railway networks in
emergencies, and the computational results of a two-step solution algorithm are
satisfied. It also shows that the method to calculate the train numbers of different
types is feasible and efficient. The genetic algorithm is successfully introduced in
train stops setting, which is very easy to understand and realize. The stop plan
designed based on the approaches in this chapter meets the demand of most of the
passengers in emergencies.
74 5 Line Planning in Emergencies for Railway Network

Future research is directed toward a generation of the model to generate the line
planning for the railway on a more complex network, not only under the situation of
emergencies but also under normal conditions. Line planning problem on a larger
scale railway network can also be described by the model presented in this chapter.
For dealing with the more complex line planning problem, extensions of the model
with modular and hierarchical could be studied and utilized in the future research
work.

References

Bussieck, M. (1998). Optimal lines in public rail transport (Ph.D. thesis). Technischen Universität
Braunschweig, Braunschweig.
Carey, M., & Lockwood, D. (1995). Model, algorithms and strategy for train pathing. Journal of
the Operational Research Society, 46(8), 988–1005.
Chakroborty, P., & Dwivedi, T. (2002). Optimal route network design for transit systems using
genetic algorithms. Engineering Optimization, 34(1), 83–100.
Ghoseiri, K., & Morshedsolouk, F. (2006). ACS-TS: Train scheduling using ant colony system.
Journal of Applied Mathematics and Decision Sciences, Article ID 95060, 28. doi:10.1155/
JAMDS/2006/95060
Guihaire, V., & Hao, J. (2008). Transit network design and scheduling: A global review.
Transportation Research Part A, 42, 1251–1273.
Jovanovic, D., & Harker, P. (1991). Tactical scheduling of rail operations: SCAN I system.
Transportation Sciences, 25(1), 46–64.
Lindner, T. (2000). Train schedule optimization in public rail transport (Ph.D. thesis).
Technischen Universität Braunschweig, Braunschweig.
Mandl, C. E. (1979). Evaluation and optimization of urban public transportation networks.
European Journal of Operational Research, 5, 396–404.
Meng, X., Jia, L., Qi, Y., Xu, J., & Wang, L. (2012). Calculation of railway transport capacity in an
emergency based on Markov process. Journal of Beijing Institute of Technology, 21(1), 77–80.
Michaelis, M., & Schöbel, A. (2009). Integrating line planning, timetabling, and vehicle
scheduling: A customer-oriented heuristic. Public Transport, 1(3), 211–232.
Murray, A. T. (2003). A coverage model for improving public transit system accessibility and
expanding access. Annals of Operations Research, 123, 143–156.
Patz, A. (1925). Die richtige Auswahl von Verkehrslinien bei gro en Stra enbahnnetzen.
Verkehrstechnik, 50/51.
Sonntag, H. (1977). Linienplanung im ffentlichen personennahverkehr (Ph.D. Thesis). Technische
Universitt Berlin, Berlin.
Xiong, Y., & Schneider, J. B. (1993). Transportation network design using a cumulative algorithm
and neural network. Transportation Research Record, 1364, 37–44.
Zhao, F., & Ubaka, I. (2004). Transit network optimization—Minimizing transfers and optimizing
route directness. Journal of Public Transportation, 7(1), 67–82.
Zhao, F., & Zeng, X. (2006). Simulated annealing–genetic algorithm for transit network
optimization. Journal of Computing in Civil Engineering, 20(1), 57–68.
Chapter 6
Train Re-Pathing in Emergencies Based
on Fuzzy Linear Programming

Abstract Train pathing is a typical problem which is to assign the train trips on the
sets of rail segments, such as rail tracks and links. This chapter focuses on the train
pathing problem, determining the paths of the train trips in emergencies. We ana-
lyze the influencing factors of train pathing, such as transferring cost, running cost,
and social adverse effect cost. With the overall consideration of the segment and
station capability constraints, we build the fuzzy linear programming model to solve
the train pathing problem. We design the fuzzy membership function to describe the
fuzzy coefficients. Furthermore, contraction–expansion factors are introduced to
contract or expand the value ranges of the fuzzy coefficients, coping with the
uncertainty of the value range of the fuzzy coefficients. We propose a method based
on triangular fuzzy coefficient and transfer the train pathing (fuzzy linear pro-
gramming model) to a determinate linear model to solve the fuzzy linear pro-
gramming problem. An emergency is presented which is based on the real data of
the Beijing–Shanghai Railway. The model in this chapter was solved and the
computation results prove the availability of the model and efficiency of the
algorithm.

6.1 Introduction

Nowadays, railway transportation needs to become more and more competitive, so

new features are required to improve the planning process. There are two
approaches to improve the capacity of the railway infrastructure. One is to enhance
the construction of railway infrastructure, such as extending the rail tracks and
improving performance of the signaling systems. The other is to utilize the existed
infrastructure more efficiently. It is generally believed that the railway operation
work can be divided into three levels, strategic level, tactical level, and operational
level (Narayanaswami and Rangaraj 2011). The strategic level is about trans-
portation pattern selecting, which is related to the national transportation policy.
And the middle one, tactical level, is on the line plan designing, which is also called
service plan, determining the trains number, paths and stops, etc.

© Springer Nature Singapore Pte Ltd. 2017 75

L. Jia et al., Train Operation in Emergencies, Advances in High-speed
Rail Technology, DOI 10.1007/978-981-10-4597-4_6
76 6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

And the line plan is divided into several parts, which are the origin and desti-
nation stations determining, the trains number calculation, the train pathing, and the
stops setting. Among them, train pathing is the most important step to design the
whole line plan, which is the basis of stops setting. Generally, the paths of the trains
are relatively steady, according to the yearly railway line plan. However, there are
occasional railway accidents which reduce the capability of the railway line and
make it impossible for the trains to run on the planned paths. It is necessary to find
the substitute path for the trains. On the other hand, with the increase of the
available rail, the topology structure of the railway network is changing profoundly.
A new railway network is forming gradually, which makes it possible that more
than one path can be found for the trains and train trips can be allocated on the
paths.
The organization of this chapter is as follows. Following this introduction, we
first discuss related works on the problem in Sect. 6.2. Then we build the train
pathing model based on fuzzy linear programming in Sect. 6.3. In Sect. 6.4, we
analyze the fuzzy coefficients in the train pathing model and design a new algorithm
to solve the fuzzy linear model. Furthermore, we study the values range of the fuzzy
coefficients, designing a method to describe the uncertainty of the fuzzy characters
of the coefficients. We prove the availability of the model and the efficiency of the
algorithm with a computation case in Sect. 6.5. In Sect. 6.6, we draw a conclusion.

6.2 Related Works

Caprara et al. (2007) and D’Ariano and Pranzo (2009) grouped the major published
railway operation as line planning, timetabling, platforming, rolling stock man-
agement, shunting, and crew planning. Train pathing is a key step of line planning,
which belongs to the tactical level. Train timetables are usually specified after the
train pathing (Cordeau et al. 1998). So, it is a must to determine the path plan before
timetabling, especially in emergencies.
There are two kinds of approaches to solve the train pathing problem in the
limited number of publications, the mathematical approaches and heuristic
approaches.
Carey (1994a) presented a mathematical model, algorithms, and strategy for
pathing trains of different speeds and stopping patterns for a double track rail line
dedicated to trains in one direction. The model included track assignment to trains
within stations (choice of platform) and between stations (choice among multiple
lines). Station layout was also considered in the model. He applied the model to a
small network and found an acceptable solution. He further extended the model
from one-way to two-way-tracks (Carey 1994b). Carey and Lockwood (1995)
developed a model and algorithm for the TPP for one train line with station stops,
and solved instances of 10 trains and 10 links. All the trains on the line travel in the
same direction. D’Ariano et al. (2008) hired a branch-and-bound algorithm for
sequencing train movements, while a local search algorithm is developed for
6.2 Related Works 77

rerouting optimization purposes. And they analyzed different types of disturbances,

including train delays and blocked tracks. The authors of this chapter defined
generating paths in emergencies as a k-shortest path problem and proposed the
method to solve it, innovating Dijkstra algorithm (Meng et al. 2010). Fuzzy pro-
gramming is introduced to solve the train routing and pathing problem recently.
And Yang et al. (2011) considered the fuzziness in the railway transportation
system, proposed a min–max chance-constrained programming model to solve the
freight train routing problem with fuzzy information.
Heuristic is also hired in train pathing problem solving in recent years. Carey and
Crawford (2007) developed a heuristic to settle for a plan that brings trains through
a rail corridor with multiple lines and multiple stations. They started from algo-
rithms that schedule trains at a single train station, and extend these to handle a
series of complex stations linked by multiple one-way lines in each direction,
traversed by trains of differing types and speeds. The algorithm was based on a set
of rules to resolve the conflicts. Lee and Chen (2009) also presented a heuristic that
includes both train pathing and train timetabling, and has the capability to solve
real-sized instances. This heuristic allowed the operation time of trains to depend on
the assigned track. Blum and Eskandarian (2002) used a delegation model to
improve agent collaboration is an effective way to improve the efficiency of an
A-Team for railroad flow optimization, including train pathing and railroad routing.
Erlebach et al. (2001) studied the method to assign trains to satisfy scheduled routes
in a cost-efficient way and proposed approximation algorithms. Törnquist (2007)
presented a heuristic approach for railway traffic rescheduling during disturbances
and a performance evaluation for various disturbance settings using data for a large
part of the Swedish railway network. Dorfman and Medanic (2004) developed a
local feedback-based travel advance strategy, using a discrete event model of train
advances along the lines of the railway to quickly handle perturbations on the
railway network, including train pathing. Caimi et al. (2011) addressed the problem
of generating conflict-free train schedules on a microscopic model of the railway
infrastructure and developed an alternative model using the sequence of resources
that each train path passes, encoded in a resource tree. They showed that the
number of maximal conflict cliques is linear in the number of train paths and
verified the model with real-world data from the Swiss Federal Railways. Lusby
et al. (2013) described a set packing inspired formulation of train routing problem
and developed a branch-and-price-based solution approach. They verified the model
with the test instance arising in Germany and supplied by the major German railway
company, Deutsche Bahn. Pellegrini et al. (2014) proposed a mixed-integer linear
programming formulation for tackling this problem, representing the infrastructure
with fine granularity. They tackled randomly generated instances representing
traffic in the control area named triangular of Gagny, and instances obtained from
the real timetable of the control area including the Lille Flandres station (both in
France) and found that negative impact of a rough granularity on the delay suffered
by trains was remarkable and statistically significant. Li et al. (2013) constructed a
train routing model combined with a train scheduling problem, which is a 0–1
mixed-integer nonlinear programming problem. They designed a Tabu search
78 6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

procedure to further improve the route schemes. Train re-pathing problem is similar
with the train routing problem in several aspects. So their approach also gave us
some enlightenment.
All these related works gave us much enlightenment when we built the train
re-pathing model and designed the algorithm to solve it. However, fuzzy charac-
teristics of train re-pathing problem were not considered in these publications, and
the rail segments capability is not set to be the restriction when building the model
in most of the publications. Therefore, we also focus on the processing of fuzzy
coefficients processing in the train re-pathing model.

6.3 Train Re-Pathing Model

The objective is to reduce the total cost as much as possible. The input data include
the paths between two stations, the capability of the rail segments affected, and the
stations affected and all the trains information needing changing paths.

6.3.1 Basic Assumption

(1) Assumption on crew. We took it for granted that the crew resource is sufficient
to cope with the trains flow distribution.
(2) Assumption on rails availability. We took it for granted that all the trains can
run on all the types of rails.

6.3.2 Graph-Based Description of Rail Networks

G ¼ ðV; EÞ is a railway network that is constructed of all kinds of rails. V is the set
of vertexes in the railway network. E is the set of edges in the railway network. V
includes the stations of the existent normal speed railway, the existent intercity
railway, and the newly built railway. And E not only includes the rail segments of
the different types of railway, but also includes the links between different types of
rails.

6.3.3 Available Paths Set Generating

According to the method in our previous research paper (Meng et al. 2009), we can
generate the paths set P when an emergency occurs. The calculating steps are as
follows.
6.3 Train Re-Pathing Model 79

Step 1: To find the shortest path with Dijkstra algorithm between the origin and the
destination. Put the shortest path, length of the shortest paths and nodes on the
shortest path into the path array P, distance array D, and node array M.
Step 2: To find neighbor nodes of the shortest path in array P and put them into
another array N.
Step 3: To calculate the distance of n-shortest path of vs-vt-vj-ve, which pass through
neighbor vt and put it into array T. vj is a node on n-shortest path.
Step 4: To order the lengths values in array T. To select the smallest one and put the
relative path in array P. To add 1 to the number of the shortest paths.
Step 5: If the total capability of the all the shortest paths reaches to the required
capability, stop the calculation. Else, go step 2.
Then we can generate a set of shortest paths for the train operation and the sum
of capability of all the paths in the path set is enough for train re-pathing work.

6.3.4 Optimization Objectives

The cost can be divided into three parts, the running cost, transferring cost, and
social effect punishment cost. The running cost is an inevitable cost, which occurs
during the running process.
When distributing the trains on paths, which consist of different kinds of rails,
the transferring cost and the social effect punishment cost occur. In this chapter,
transferring cost is used to denote the cost occurs when a train transfers from one
type of rail line to another type of rail line. Transferring cost includes equipment
cost, technology operation cost, and abrasion cost (Pu 1999). Among them the
equipment costs and abrasion costs are very difficult to calculate accurately. The
technology operation cost is related to profit of the railway bureau and the tech-
nology operation quantity. The transferring cost also depends on the rail grade,
train type, and the fact whether a ferry-locomotive is needed, which is very difficult
to calculate exactly. But we can set the value range of it.
The social effect punishment cost is related to the passenger satisfaction, which
is also difficult to figure out and the value range can be defined.
Transferring cost and social effect punishment cost are more characterized by
fuzziness in actual transportation operation, especially in emergencies. The coef-
ficient can be expressed by some fuzzy functions, such as triangular fuzzy function
and trapezoidal fuzzy function. All the optimization objectives can be compromised
to some extent. As long as the values of the optimization objectives reach into a
certain value range, it is considered that the optimization process is successful. We
designed the method to cope with the fuzzy character of all the objectives and the
algorithm to solve the trains flow distribution problem.
80 6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

6.3.5 Train Distribution Model

The decision variables and parameters are as follows.

V the set of all the station nodes in the network considered in this chapter
k index of train type
M the number of the train types
p index of path
P the set of available p
Qp the set of all segments on path p
Rp the set of all stations on path p
s index of station
t index of station
vs , vt station node s and station node t
est the segment from station vs to station vt
dts the length from station vs to station vt
nkp the number of the k type trains allocated at path p
dst the transferring cost coefficient from station vs to station vt
nst the running cost coefficient of segment from station vs to station vt
kkp the social cost punishment coefficient if a k type train allocated at path p
Dst the capability of segment est
Bs the capability of station s
Nþ the set of positive integers

6.3.5.1 Formulation of Objectives

The costs are listed and analyzed in Sect. 6.3.4. Here we formulate the costs
respectively.
(1) Transferring cost

M X X
X
ZT ¼ dst nkp ð6:1Þ
k¼1 p2P est 2Qp

(2) Running cost

!
M X
X X
ZR ¼ nkp nst dts ð6:2Þ
k¼1 p2P est 2Qp
6.3 Train Re-Pathing Model 81

(3) Social effect punishment cost

M X
X
ZS ¼ kkp nkp ð6:3Þ
k¼1 p2P

Then, we normalized the three kinds of cost by adding the coefficients, dst , nkt and
kkp . Then the total cost of the model is as follows.

min ZTT ¼ ZT þ ZR þ ZS ð6:4Þ

It is equal to
!
M X X
X M X
X X X
M X
min ZTT ¼ dst nkp þ nkp nst dts þ kkp nkp ð6:5Þ
k¼1 p2P est 2Qp k¼1 p2P est 2Qp k¼1 p2P

6.3.5.2 Constraints of the Model

There are many constraints when assigning all the trains to the available paths. The
main constraints to be considered are as follows.
(1) Segments capacity constraints
The number of trains running through segment est cannot surpass its capability.

X
M X X
nkp  Dst ð6:6Þ
k¼1 p2P est 2Qp

Dst is the capability of segment est .

(2) Stations capacity constraints
The capability of every station in the railway network is bigger than the number
of all the trains inbound and outbound.

X
M X X
nkp  Bs ð6:7Þ
k¼1 p2P vs 2Rp
82 6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

(3) Non-negativity constraints

nkp 2 N þ or nkp ¼ 0 ð6:8Þ

We can see that the model is a linear integer programming model.

6.4 Fuzzy Coefficients Processing and Train Re-Pathing

Model Solution

There are numerous fuzzy numbers in the model built up in Sect. 6.3.5. Thus, we
first present the method to process fuzzy numbers of the model. Then based on the
processing, we propose the steps to set the model with optimization software
LINGO 11.0.

6.4.1 Fuzzy Coefficients Processing

A fuzzy number is a generalization of a regular, real number in the sense that it does
not refer to one single value but rather to a connected set of possible values, where
each possible value has its own weight between 0 and 1. This weight is called the
membership function. In the engineering computation field, many elements cannot
be described with definite numbers, while we can tell how much they belong to a
certain range. The degree can be represented by fuzzy numbers. It is a powerful tool
to describe this kind of element.
Generally, fuzzy linear programming models can be divided into three groups.
The first group of models has fuzzy resources in the constraints of the model. That
is to say, the resources of the constraints are fuzzy which should be described with
the fuzzy membership functions. The second group of models has the fuzzy
coefficients of the objectives. The fuzzy numbers occur in the optimization goal
equations. The last group has the characteristics of the above two groups. They both
have the fuzzy resources constraints and the fuzzy objective coefficients.
In this chapter, there are several objective coefficients which are uncertain and
difficult to obtain and we model the problem as the second group. Transferring cost is a
typical fuzzy number and it is very difficult to get. When disturbs occur, the price
assessment of transferring cost is with more fuzziness. Fuzzy factors could be defined
with fuzzy numbers. Typical fuzzy membership functions are triangular function,
trapezoid function, and so on. When the fuzzy degree is out control with the typical
definition of the fuzzy factors, we should improve the function to deal with the situation.
It is clear that the train distribution model is a fuzzy linear integer programming
model. The tolerance method is the most typical method. In this section, we
introduce the tolerance method and present a new method to solve the fuzzy linear
6.4 Fuzzy Coefficients Processing and Train Re-Pathing Model Solution 83

integer programming model. And we propose a method to enlarge the fuzzy

coefficients support.
In some occasions, the boundaries of the value range are also difficult to
determine, especially in emergencies. So, we design a method, hiring a function
FðxÞ to expand the value range.
Set EH to be the optimistic value of d, and EG to be the pessimistic value. Then
EG  d  EH , EG  0; EH  0,
d is the average value of d, as shown in Fig. 6.1a.

0 EB EB + ET ET δ
2

(a) Original range of the fuzzy variables

0 EG EG + EH EH
δ
2

(b) The longitudinal axis is been moved to the right

EH − EG 0 EH − EG x
−
2 2

(c) Original range of the fuzzy variables after the longitudinal axis is moved

EH − EG EH − EG
− F ( x) F ( x)
2 2

EH − EG EH − EG 0 EH − EG EH − EG x
−U − U
2 2 2 2

(d) Alterable range of the fuzzy variables

Fig. 6.1 Design of alterable range of the fuzzy variables

84 6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

Set

x ¼ d  ðEG þ EH Þ=2 ð6:9Þ

The change is shown in Fig. 6.1b.

Then

ðEH  EG Þ=2  x  ðEH  EG Þ=2 ð6:10Þ

So x is symmetrical by y axis, as shown in Fig. 6.1c.

Then, FðxÞ is hired to expand the value range of x.

EH  EG EH  EG
FðxÞ  x  FðxÞ ð6:11Þ
2 2

So

EH  EG EG þ EH EH  EG EG þ EH
FðxÞ þ  d  FðxÞ þ ð6:12Þ
2 2 2 2

It can be seen in Fig. 6.1d.

That is to say,

EH  EG EG þ EH
 Fðd  ðEG þ EH Þ=2Þ þ
2 2
EH  EG EG þ EH
 d  Fðd  ðEG þ EH Þ=2Þ þ ð6:13Þ
2 2

So the fuzzy coefficients value range is as flows after steps above.

 
EH  EG EG þ EH EH  EG EG þ EH
Fðd  ðEG þ EH Þ=2Þ þ ; Fðd  ðEG þ EH Þ=2Þ þ
2 2 2 2
ð6:14Þ

It can be seen that the expanded value range is related to the original range and
the average value of the fuzzy coefficients. This method can deal with the fuzzy
coefficients flexibly, making the coefficients close to the real cost as much as
possible.

6.4.2 Steps to Solve Train Distribution Model

It is obvious that the programming model is a fuzzy linear programming with fuzzy
objective coefficients. Since some coefficients of the objective are fuzzy, we must
deal with them first. We design the method to express the coefficients with the
6.4 Fuzzy Coefficients Processing and Train Re-Pathing Model Solution 85

pessimistic value, average value, and optimistic value. Since EH is the optimistic
value of d, and EG is the pessimistic value of d, we set EA to be the average value of
d. We assume that d ¼ w1 EH þ w2 EG þ ð1  w1  w2 ÞEA , where w1 and w2 are the
weights of the optimistic value and pessimistic value respectively. We can see that
the fuzzy linear programming can be transferred into different deterministic linear
programming with different pairs of w1 and w2 . Then the steps to solve the problem
are as follows.
Step 1: Set w2 ¼ 0:1;
Step 2: Set w2 ¼ w2 þ 0:1;
Step 3: Set w1 to be 0.1, then solve the linear programming with LINGO 11.0;
Step 4: Repeat Step 2 with w1 increasing 0.1 a time;
Step 5: Record the value of w1 , w2 and the computing results;
Step 6: Go to Step 2 and repeat the process until w2 ¼ 1:5.
Step 7: Select the satisfying solution for the model.

6.5 Case Study

6.5.1 Case Scenario

It is assumed that there is an emergency at DK856+321 on the Beijing–Shanghai

high-speed railway. Then the trains cannot run through the segment of East Xuzhou
to Bengbu. And the time required to recover is 4–8 h. The railway network around
the emergency place is shown in Fig. 6.2. We will study the trains flow distribution
problem on the downgoing direction.

6.5.2 Trains to Be Re-Pathing

Trains arrive at Xuzhou joint from 8 to 12 are as follows.

(1) The high-speed trains (shorthand: H): G301, G303, G305, G101, G103, G105,
G107, G109, G111, G113
(2) Multiple Units (shorthand: M): D88/5
(3) T trains and K trains (shorthand: T&M): K58/5, K518/5, K101/4/1
(4) Normal speed trains (shorthand: N): 1230/27
(5) Low-speed trains (shorthand: L): 10135, 10625, 11301, 23005, 11305
(6) Temporary trains (shorthand: T): None
(7) Other trains (shorthand: O): None
86 6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

203 146 109

Shangqiu Xuzhou Xinyi
Zhengzhou
168 337
Yangzhou

15
16
140 Haian
Fuyang

5
5
225
118
Luohe 126 Bengbu 100
175 101

86
162 181 Jiangyin
Nanjign
6

Shuijiahu
11

184 48
Xinyang 95 175
Huangchuan
104 166 Wuxi
126
120 Hefei 125
140
Macheng

95
234 141 Wuhu
Shanghai

8
178 126

18
125
327 Changxing
22

Hangzhou
3

Wuhan 20
262 Xiaoshan 145
Ningbo
Jiujiang

Fig. 6.2 Railway network around the emergency place. Note The broader lines stand for
high-speed rails and the thin lines stand for low speed rails

6.5.3 Available Paths

According to the method in our previous paper (Meng et al. 2009), we generate the
available paths according to the succinct description in Sect. 6.3.3, shown in
Table 6.1 and Fig. 6.3.

Table 6.1 Available paths No. Available paths Length (km)

from Xuzhou to Nanjing
(by C-enough plan) 1 1-2-5-6-3 336
2 1-4-5-6-3 346
3 1-4-5-7-8-6-3 502
And lengths of every segments on the paths are shown in
Table 6.2
6.5 Case Study 87

1 2 3
(3)

4 5 6

7 8

Xuzhou Bengbu Shuijiahu Hefei Nanjing

Fig. 6.3 Paths between Xuzhou and Nanjing in an emergency

Table 6.2 Lengths of Segment No. Segments Length (km)

segments in the paths
1 1–2 155
2 4–5 165
3 5–6 181
4 5–7 86
5 7–8 95
6 8–6 166

6.5.4 Distribution Plan

6.5.4.1 Specification of the Train Distribution Model

We can see that there are three available paths in the partial railway network, which
are marked (1), (2) and (3), as shown in Fig. 6.3. Eight stations and six segments
are in the network.
Now the goal is to allocate the trains on the three paths in Table 6.1.
We specified the model as follows:

minZ ¼ d14 ðn12 þ n13 Þ þ d25 ðn11 Þ þ d63 ðn11 þ n12 þ n13 Þ
þ n12 n11  155
þ n45 ðn12 þ n22 þ n32 þ n42 þ n52 þ n13 þ n23 þ n33 þ n43 þ n53 Þ  165
þ n56 ðn11 þ n12 þ n22 þ n32 þ n42 þ n52 Þ  181
þ n57 ðn13 þ n23 þ n33 þ n43 þ n53 Þ  86
þ n78 ðn13 þ n23 þ n33 þ n43 þ n53 Þ  95
þ n86 ðn13 þ n23 þ n33 þ n43 þ n53 Þ  166
þ k11 ðn11 Þ þ k12 ðn12 Þ þ k13 ðn13 Þ
þ k23 n23 þ k33 n33 þ k43 n43 þ k53 n53

s.t.
88 6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

8 1
>
> n1  C21
>
>
>
> n12 þ n13 þ n22 þ n23 þ n32 þ n33 þ n42 þ n43 þ n52 þ n53  C54
>
>
>
> n11 þ n12 þ n22 þ n32 þ n42 þ n52  C65
>
>
>
> n13 þ n23 þ n33 þ n43 þ n53  C75
>
>
>
> n3 þ n3 þ n3 þ n3 þ n3  C 8
1 2 3 4 5 7
>
>
>
> n3 þ n3 þ n3 þ n3 þ n3  C68
1 2 3 4 5
>
>
>
> n11  B2
>
>
> n12 þ n13 þ n22 þ n23 þ n32 þ n33 þ n42 þ n43 þ n52 þ n53  B4
>
>
< 1
n1 þ n12 þ n13 þ n22 þ n23 þ n32 þ n33 þ n42 þ n43 þ n52 þ n53  B5
>
> n11 þ n12 þ n13 þ n22 þ n23 þ n32 þ n33 þ n42 þ n43 þ n52 þ n53  B6
>
>
>
> n13 þ n23 þ n33 þ n43 þ n53  B7
>
>
>
> n13 þ n23 þ n33 þ n43 þ n53  B8
>
>
>
> n11 þ n12 þ n13 ¼ 10
>
>
>
> n21 þ n22 þ n23 ¼ 1
>
>
>
> n31 þ n32 þ n33 ¼ 3
>
>
> n41 þ n42 þ n43 ¼ 1
>
>
>
> n51 þ n52 þ n53 ¼ 5
>
>
: nk ¼ 0 or nk 2 N þ ; k ¼ 1; 2; 3; 4; 5; p ¼ 1; 2; 3
p p

d14 , d25 , d63 are got from the publication (Pu 1999).
k23 , k33 , k43 , k53 are social effect punishment cost coefficient when the 2th, 3th, 4th,
5th kinds of trains are allocated on path 3. The can be attained by the Delphi
method.
We got the train running cost coefficients according to the data listed in two
publications (Li and Lu 1997; Qi and Xiong 2008). The coefficients are as follows.

n12 ¼ n86 ¼ 325; n45 ¼ n56 ¼ n57 ¼ n78 ¼ 92:4:

The capabilities of the related segments and stations are as follows.

C21 ¼ 115; C54 ¼ 20; C65 ¼ 20; C75 ¼ 12; C87 ¼ 12; C68 ¼ 62; B2 ¼ 30; B4 ¼ 32
B5 ¼ 38; B6 ¼ 30; B7 ¼ 20; EG B8 ¼ 70

Set d14 * (2800, 3000, 3200), d25 * (2600, 2800, 3000), d63 * (3000, 3200,
3400), k11 * (20,000, 22,000, 24,000), k12 * (60,000, 64,000, 68,000),
k13 * (80,000, 86,000, 92,000), k23 , k33 , k43 , k53 * (10,000, 11,000, 12,000).
For example, d14 * (2800, 3000, 3200) means that the largest value of the fuzzy
number d14 is 3200 and the smallest value is 2800. The average value is 3000. That
is to say that EH ¼ 3200 and EG ¼ 2800.
If at this time w1 ¼ 0:1 and w2 ¼ 0:5, d14 ¼ 0:1  3200 þ 0:5  2800 þ
ð1  0:1  0:5Þ  3000 ¼ 2920. All the other fuzzy coefficients can be calculated
out in the same way. The fuzzy linear programming model is turned into a
6.5 Case Study 89

deterministic linear programming model, which can be easily solved with the
software LINGO 11.0.
It should be noticed that there is little transferring cost at Hefei and Nanjing on
path (3), for the segment between Hefei and Nanjing is a high-speed segment. But
the transferring operation is in the station, and the cost is very little. So, this
transferring cost is not taken into consideration in this model.

6.5.4.2 Solutions

(1) Solving the problem in original value range

We compute the results respectively while w1 is assigned to be 0.1–1.5. The
results are shown in Table 6.3.
It can be seen that the fuzzy coefficients are bigger than the average value. It
means that the smaller the value of the fuzzy coefficients, the bigger of the objective
value is. It is obvious that the relative results are not satisfying. So the results in
shadowed part in Table 6.3 are the unreasonable solution.
When w1 is 0.5, the fuzzy coefficients are equal to their average value. This is the
most possible situation of the reality, in which the fuzzy membership is 1. The total
cost is 1.2778E06 at this point, which is the highest.
When w1 is set to be 0.7–1.5, solution of the problem is not changed. That is to
say, all the solutions with the membership under 0.8 are the same. However, the
objective value changes with the fuzzy coefficients changing. According to the rule
that the solution with maximal membership value and the minimum objective value
should be selected, the solution is taken as n11 ¼ 10, n12 ¼ 0, n22 ¼ 1, n32 ¼ 3, n42 ¼ 1,
n52 ¼ 5, n13 ¼n23 ¼ n33 ¼n43 ¼ n53 ¼ 0. The objective value is 1.2659E06.
(2) Solving the problem in variable value range
Set the variable function FðxÞ ¼ 2. So, the value range is twice as large as the
original value range. That is to say, the expanding coefficient is 2. We take the
calculation of d14 for instance. Since FðxÞ ¼ 2, EG = 2800 and EH = 3200, the value
range of d14 is
 
EH  EG EG þ EH EH  EG EG þ EH
FðxÞ þ ; FðxÞ þ
2 2 2 2
 
3200  2800 3200 þ 2800 3200  2800 3200 þ 2800
¼ 2  þ ;2  þ
2 2 2 2
¼ ½2600; 3400:

All the other coefficients value range can be obtained by the same means. In fact,
we expand the value range by this means based on the original value range,
according to Eq. (6.12).
 90

Table 6.3 Computation results of the trains flow distribution model

w1 F1 F2 F3 F4 F5 F6 F7 n11 n12 n13 n22 n23 n32 n33 n42 n43 n52 n53 Smf Z (10^6)
(a) In original value range
0.1 3080 2880 3280 22,800 65,600 88,400 11,400 9 1 0 1 0 2 1 1 0 3 2 4.20 1.3684
0.2 3060 2860 3260 22,600 65,200 87,800 11,300 9 1 0 1 0 2 1 1 0 4 1 4.90 1.3389
0.3 3040 2840 3240 22,400 64,800 87,200 11,200 9 1 0 1 0 2 1 1 0 4 1 5.60 1.3361
0.4 3020 2820 3220 22200 64,400 86,600 11,100 9 1 0 1 0 3 0 1 0 4 1 6.30 1.3068
0.5 3000 2800 3200 22,000 64,000 86,000 11,000 9 1 0 1 0 3 0 1 0 5 0 7.00 1.2778
0.6 2980 2780 3180 21,800 63,600 85,400 10,900 9 1 0 1 0 3 0 1 0 5 0 6.30 1.2752
0.7 2960 2760 3160 21,600 63,200 84,800 10,800 10 0 0 1 0 3 0 1 0 5 0 5.60 1.2659
0.8 2940 2740 3140 21,400 62,800 84,200 10,700 10 0 0 1 0 3 0 1 0 5 0 4.90 1.2635
0.9 2920 2720 3120 21,200 62,400 83,600 10,600 10 0 0 1 0 3 0 1 0 5 0 4.20 1.2611
1.0 2900 2700 3100 21,000 62,000 83,000 10,500 10 0 0 1 0 3 0 1 0 5 0 3.50 1.2587
1.1 2880 2680 3080 20,800 61,600 82,400 10,400 10 0 0 1 0 3 0 1 0 5 0 2.80 1.2563
1.2 2860 2660 3060 20,600 61,200 81,800 10,300 10 0 0 1 0 3 0 1 0 5 0 2.10 1.2539
1.3 2840 2640 3040 20,400 60,800 81,200 10,200 10 0 0 1 0 3 0 1 0 5 0 1.60 1.2515
1.4 2820 2620 3020 20,200 60,400 80,600 10,100 10 0 0 1 0 3 0 1 0 5 0 0.70 1.2491
1.5 2800 2600 3000 20,000 60,000 80,000 10,000 10 0 0 1 0 3 0 1 0 5 0 0.00 1.2467
(b) In variable value range
0.1 3160 2960 3360 23,600 66,400 89,200 12,200 9 1 0 1 0 2 1 1 0 3 2 4.20 1.3804
0.2 3120 2920 3320 23,200 65,800 88,400 11,900 9 1 0 1 0 2 1 1 0 4 1 4.90 1.3473
0.3 3080 2880 3280 22,800 65,200 87,600 11,600 9 1 0 1 0 2 1 1 0 4 1 5.60 1.3412
0.4 3040 2840 3240 22,400 64,600 86,800 11,300 9 1 0 1 0 3 0 1 0 4 1 6.30 1.3094
0.5 3000 2800 3200 22,000 64,000 86,000 11,000 9 1 0 1 0 3 0 1 0 5 0 7.00 1.2778
0.6 2960 2760 3160 21,600 63,400 85,200 10,700 9 1 0 1 0 3 0 1 0 5 0 6.30 1.2728
0.7 2920 2720 3120 21,200 62,800 84,400 10,400 10 0 0 1 0 3 0 1 0 5 0 5.60 1.2611
6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

(continued)
Table 6.3 (continued)
w1 F1 F2 F3 F4 F5 F6 F7 n11 n12 n13 n22 n23 n32 n33 n42 n43 n52 n53 Smf Z (10^6)
0.8 2880 2680 3080 20,800 62,200 83,600 10,100 10 0 0 1 0 3 0 1 0 5 0 4.90 1.2563
0.9 2840 2640 3040 20,400 61,600 82,800 9800 10 0 0 1 0 3 0 1 0 5 0 4.20 1.2515
6.5 Case Study

1.0 2800 2600 3000 20,000 61,000 82,000 9500 10 0 0 1 0 3 0 1 0 5 0 3.50 1.2467
1.1 2760 2560 2960 19,600 60,400 81,200 9200 10 0 0 1 0 3 0 1 0 5 0 2.80 1.2419
1.2 2720 2520 2920 19,200 59,800 80,400 8900 10 0 0 1 0 3 0 1 0 5 0 2.10 1.2371
1.3 2680 2480 2880 18,800 59,200 79,600 8600 10 0 0 1 0 3 0 1 0 5 0 1.60 1.2323
1.4 2640 2440 2840 18,400 58,600 78,800 8300 10 0 0 1 0 3 0 1 0 5 0 0.70 1.2275
1.5 2600 2400 2800 18,000 58,000 78,000 8000 10 0 0 1 0 3 0 1 0 5 0 0.00 1.2227
The solutions in bold are the unreasonable solutions
91
92 6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

The expanded value ranges are as follows.

d14  ð2600; 3000; 3400Þ; d25  ð2400; 2800; 3200Þ; d63  ð2800; 3200; 3600Þ;
k11  ð18; 000; 22; 000; 26; 000Þ; k12  ð58; 000; 64; 000; 70; 000Þ;
k13  ð78; 000; 86; 000; 94; 000Þ; k23 ; k33 ; k43 ; k53  ð8000; 11; 000; 14; 000Þ:

The computation results are shown in Table 6.3.

From Table 6.3 we can see that the objective value is reduced when the mem-
bership values are the same, as shown in Fig. 6.4.
The difference between the objective value calculated with the original value
range and that calculated with the variable value range becomes obvious from the
point where w1 is 0.7. And the difference is becoming more and more obvious till
the point where w1 is 1.5. The objective value is 1.2467E06 with the original value
range and the objective value is 1.2227E06 with the variable value range. So we can
see the object value is optimized when we make the value range variable.
The solution with the variable value range is also n11 ¼ 10, n12 ¼ 0; n22 ¼ 1,
n2 ¼ 3; n42 ¼ 1, n52 ¼ 5, n13 ¼ n23 ¼ n33 ¼ n43 ¼ n53 ¼ 0.
3

And we can see in Fig. 6.5 that all the trains are allocated on path 1 and path 2.
No train is allocated on path 3. It is related to the required capacity which is 40,
when searching for the available paths. However, the number of trains needing to be
allocated is 20. It is necessary to set the required capacity to be bigger than the
number of trains needing to be allocated. For one thing, the accurate number of the
trains needing to be allocated is difficult to forecast. For another, it is a must to
reserve extra capacity to deal with the uncertain situation of the reality.
The methods presented in this chapter can give the optimized solution, satisfying
the fuzzy membership constraint. So, we can deal with the fuzzy character of the
trains flow distribution model to approach the reality as possible as we can. We can

Fig. 6.4 Solution of the train 1.29

paths distributing model Original value range
Variable value range
1.28
Objective value (×10 6 )

1.27

1.26

1.25

1.24

1.23

1.22
0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5
w1
6.5 Case Study 93

1 2 3
(3)
H:10
4 5 6
M:1,T&K:3,N:1,L:5
7 8

Xuzhou Bengbu Shuijiahu Hefei Nanjing

Fig. 6.5 Solution of the train paths distributing model shown on the railway network graph

propose several available solutions, at different fuzzy membership level for the
managers to make the decision.

6.6 Brief Summary

This chapter proposes a feasible, effective approach to solve the fuzzy programming
problems in railway transportation. We first present a model for distributing trains
on paths, offering the theory basis for train dispatching on Chinese railway network.
Then we integrate the transferring cost, running cost, and social effect punishment
cost to design the objective of the train distribution model. The character of the
coefficient of the costs is described with fuzzy membership function. And we
present a method to expand the fuzzy number value range, supporting the algorithm
to solve the model. A triangular membership function is designed to turn the fuzzy
programming model into definitive programming problem. And the detailed steps
to solve the model are given.
The method presented can also be used to solve other problems in railway
transportation organization. We can deal with the fuzzy character of the passenger
transportation and freight transportation requirement in service planning. It also
may work in fuzzy objectives in the Electric Multiple Units timetable designing, the
work time in crew schedule designing. And in solving the routing problem of trains
at stations, we can also hire the method to describe the fuzzy character when the
operation time has the fuzzy characters.

References

Blum, J., & Eskandarian, A. (2002). Enhancing intelligent agent collaboration for flow
optimization of railroad traffic. Transportation Research Part A, 36(10), 919–930.
Caimi, G., Chudak, F., Fuchsberger, M., Laumanns, M., & Zenklusen, R. (2011). A new
resource-constrained multicommodity flow model for conflict-free train routing and schedul-
ing. Transportation Science, 45(2), 212–227.
Caprara, A., Kroon, L. G., Monaci, M., Peeters, M., & Toth, P. (2007). Passenger railway
optimization, Chap. 3. In C. Barnhart & G. Laporte (Eds.), Handbooks in operations research
and management science (Vol. 14, pp. 129–187). Amsterdam: Elsevier.
94 6 Train Re-Pathing in Emergencies Based on Fuzzy Linear Programming

Carey, M. (1994a). A model and strategy for train pathing with choice of lines, platforms, and
routes. Transportation Research Part B, 28(5), 333–353.
Carey, M. (1994b). Extending a train pathing model from one-way to two-way track.
Transportation Research Part B, 28(5), 395–400.
Carey, M., & Crawford, I. (2007). Scheduling trains on a network of busy complex stations.
Transportation Research Part B, 41(2), 159–178.
Carey, M., & Lockwood, D. (1995). Model, algorithms and strategy for train pathing. Journal of
the Operational Research Society, 46(8), 988–1005.
Cordeau, J., Toth, P., & Vigo, D. (1998). A survey of optimization models for train routing and
scheduling. Transportation Science, 32(4), 380–404.
D’Ariano, A., Corman, F., Pacciarelli, D., & Pranzo, M. (2008). Reordering and local rerouting
strategies to manage train traffic in real time. Transportation Science, 42(4), 405–419.
D’Ariano, A., & Pranzo, M. (2009). An advanced real-time train dispatching system for
minimizing the propagation of delays in a dispatching area under severe disturbances.
Networks and Spatial Economics, 9(1), 63–84.
Dorfman, M. J., & Medanic, J. (2004). Scheduling trains on a railway network using a discrete
event model of railway traffic. Transportation Research Part B, 38(1), 81–98.
Erlebach, T., Gantenbein, M., Hürlimann, D., Neyer, G., Pagourtzis, A., Penna, P., et al. (2001).
On the complexity of train assignment problems. In Proceedings of ISAAC01 (Vol. 2223,
pp. 390–402). Lecture Notes in Computer Science.
Lee, Y., & Chen, C. Y. (2009). A heuristic for the train pathing and timetabling problem.
Transportation Research Part B, 43(8–9), 837–851.
Li, F., Gao, Z., Li, K., & Wang, Z. (2013). Train routing model and algorithm combined with train
scheduling. Journal of Transportation Engineering, 139(1), 81–91. (in Chinese).
Li, W., & Lu, W. (1997). Exploration on index calculating method of railway transportation cost.
Journal of the China Railway Society, 19(5), 15–20. (in Chinese).
Lusby, R. M., Larsen, J., & Ryan, D. M. (2013). A set packing inspired method for real-time
junction train routing. Computers & Operations Research, 40(3), 713–724.
Meng, X., Jia, L., Chen, C., Xu, J., Wang, L., & Xie, J. (2009). Paths generating in emergency on
china new railway network. Journal of Beijing Institute of Technology, 19(S2), 84–88.
(in Chinese).
Meng, X., Jia, L., & Qin, Y. (2010). Train timetable optimizing and re-scheduling based on improved
particle swarm algorithm. Journal of Transportation Research Record, 2(2197), 71–79.
Narayanaswami, S., & Rangaraj, N. (2011). Scheduling and re-scheduling of railway operations: A
review and expository analysis. Technology Operation Management, 2(2), 102–122.
Pellegrini, P., Marlière, G., & Rodriguez, J. (2014). Optimal train routing and scheduling for
managing traffic perturbations in complex junctions. Transportation Research Part B, 59(59),
58–80.
Pu Z. (1999). On calculation of cost of normal speed train transferring to high-speed railway
(Master’s thesis). Southweat Jiaotong University, Chengdu, China.
Qi, X., & Xiong, J. (2008). Optimization model of organization of passenger trains running
crossing PDLs and existing railways. Railway Transport and Economy, 30(4), 16–19.
Törnquist, J. (2007). Railway traffic disturbance management—An experimental analysis of
disturbance complexity, management objectives and limitations in planning horizon.
Transportation Research Part A, 41(3), 249–266.
Yang, L., Gao, Z., Li, X., & Li, K. (2011). A weighted min–max model for balanced freight train
routing problem with fuzzy information. Engineering Optimization, 43(12), 1289–1309.
Chapter 7
Train Re-scheduling Based
on an Improved Fuzzy Linear
Programming Model

Abstract Train re-scheduling remains a long-standing challenge in railway oper-

ation. To design high-quality timetable in fuzzy environment, this chapter studies
train re-scheduling problem under the fuzzy environment, in which the fuzzy
coefficients of the constraint resources have the fuzzy boundaries. Based on the
improved fuzzy linear programming, the train re-scheduling model is constructed.
Aiming at dealing with the fuzzy characteristics of the constraint coefficients value
range boundaries, the description method of this kind of objective function is
proposed and the solving approach is presented. The model has more adaptability to
model a common train re-scheduling problem, in which some resources of the
constraints are uncertain and have the characteristics of fuzziness and the bound-
aries of the resources are fuzzy. Two numerical examples are carried out and it
shows that the model proposed in this chapter can describe the train re-scheduling
problem precisely, dealing with the fuzzy boundaries of the fuzzy coefficients of the
constraint resources. The algorithm present is suitable to solve the problem. The
approach proposed in this chapter can be a reference for developers of railway
dispatching system.

7.1 Introduction on Train Re-scheduling

Railways are typically operated according to a planned (predetermined) timetable,

which determines the amount of trains and the dwell on the railway line. However,
railway accidents and natural disasters often affect the train operation, which makes
it a must to re-schedule the trains, through adjusting the inbound and outbound time
of the trains at the stations. So it is seriously important to study of train
re-scheduling problem.
The survey of Cordeau et al. reviewed a large number of papers dealing with
different problems arising in timetable design and train re-scheduling (Cordeau
et al. 1998). Narayanaswami and Rangaraj (2011) also presented a review on
scheduling and re-scheduling of railway operations, which classified the railway

© Springer Nature Singapore Pte Ltd. 2017 95

L. Jia et al., Train Operation in Emergencies, Advances in High-speed
Rail Technology, DOI 10.1007/978-981-10-4597-4_7
96 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

operations into four levels: strategic, tactical, operational control, and real-time
control. The train re-scheduling is taken as the operational control level problem.
In view of their extensive survey, we limit our review to recent papers dealing
with train re-scheduling problems. Şahin studied the real-time conflict resolution
problem on a single-track railway. Conflicts between trains are resolved in the order
in which they appear. An algorithm based on look-ahead strategies predicted
potential consecutive delays and takes ordering decisions of merging or crossing
points. The problem was formulated as a job shop scheduling problem, and the
objective is to minimize average consecutive delays (Şahin 1999). Schobel (2001)
proposed an approach which aimed to decide which connections had to be main-
tained or canceled to minimize the inconvenience for the passengers. Dorfman and
Medanic (2004) proposed a discrete-event model for scheduling trains on a single
line and a greedy strategy to obtain suboptimal schedules. The model behavior was
similar to that of human dispatchers. The authors showed that adding nonlocal
information can prevent deadlocks. The approach could quickly handle timetable
perturbations and performs satisfactorily on three time-preference criteria.
Tőrnquist and Persson (2007) discussed how disturbances propagate and which
actions to take in order to minimize the consequences for multiple stakeholders.
They presented an optimization approach to the problem of re-scheduling railway
traffic in an N-tracked network when a disturbance has occurred. Computational
results from experiments using data from the Swedish railway traffic system are
presented along with a discussion about theoretical and practical strengths and
limitations. They came to the conclusion that there is a relation between certain
disturbance characteristics and the ability to find appropriate solutions sufficiently
fast, which can be utilized to configure and improve the suggested approach further.
Chang and Kwan (2005) described the application of evolutionary computation
techniques to a real-world complex train schedule multi-objective problem. They
proposed three established algorithms (Genetic Algorithm GA, Particle Swarm
Optimization PSO, and Differential Evolution DE) to solve the scheduling problem.
They drew a conclusion that DE is the best approach for this scheduling problem.
D’Ariano et al. (2007) viewed the train scheduling problem as a huge job shop
scheduling problem with no-store constraints. They utilized a careful estimation of
time separation between trains, and described the scheduling problem with an
alternative graph formulation. They developed a branch and bound algorithm,
which included implication rules enabling to speed up the computation. Kroon et al.
(2009) generated several timetables utilizing sophisticated operations research
techniques and utilized innovative operations research tools to devise efficient
schedules for rolling stock and crew resources. They provided a new method to
generate train timetables, taking rolling-stock and crew into consideration. Kroon
et al. (1997) proved the NP-completeness of the general problem of routing trains
through railway stations to design a conflict-free timetable and show solvable
special cases.
There are also some publications on the real-time re-scheduling problem.
Mazzarello and Ottaviani (2007) described the architecture of a real-time traffic
management system that had been implemented within the European project
7.1 Introduction on Train Re-scheduling 97

COMBINE to test the feasibility of a completely automated system for conflict

resolution and speed regulation. Rodriguez (2007) proposed a heuristic approach to
train routing problems and consequent train reordering problems with operational
purposes. The algorithm was tested on a complex rail junction and can provide a
satisfactory solution within three minutes of computation time for instances
involving up to 24 trains. Adenso-Díaz et al. (1999) considered the problem of
managing real-time timetable disturbances for a regional network. They proposed
an automated conflict resolution system for the Spanish National Railway Company
and a mixed-integer programming model was adopted to describe the problem.
It is easy to see from the literature that most of researches have been carried out
under the specified environment, in which all the parameters involved are fixed
quantities. Actually, since the railway transportation system is complex, dispatchers
inevitably meet uncertain parameters when re-scheduling trains on the dispatching
sections, such as random parameters and fuzzy parameters. However, some
researchers ignored the existence of the uncertainty in the literatures, which prob-
ably caused poor quality of the re-scheduled timetable in the real applications.
Yang et al. (2011) studied the railway freight transportation planning problem
under the mixed uncertain environment of fuzziness and randomness based on the
optimization methods under the uncertain environments. They proposed a hybrid
algorithm integrating simulation algorithm and a genetic algorithm, to find optimal
paths, the amount of commodities passing through each path and the frequency of
services. It was a typical publication in which the mixed uncertain environment of
fuzziness and randomness was taken into consideration in railway operation.
Acuna-Agost et al. (2011a, b) investigated the solution of train re-scheduling
problem through a mixed-integer programming (MIP) formulation. They proposed
an approach called SAPI (Statistical Analysis of Propagation of Incidents) to limit
the search space around the original nondisrupted schedule by hard and soft fixing
of integer variables with local-branching-type cuts and proved the model effec-
tiveness with the computation cases on the railway networks of France and Chile.
Krasemann (2012) developed a greedy algorithm which performs a depth-first
search using an evaluation function when conflicts arise and then branches
according to a set of criteria to solve the train re-scheduling problem. Dündar and
Şahin (2012) developed artificial neural networks (ANNs) to mimic the decision
behavior of train dispatchers so as to reproduce their conflict resolutions.
Castillo et al. (2011) dealt with the timetabling problem of a mixed multiple- and
single-tracked railway network. Min et al. (2011) proposed a column-generation-
based algorithm that exploits the reparability of the problem. Cacchiani et al. (2010)
studied the problem of freight transportation in railway networks, where both
passenger and freight trains are run. Almodóvar and García-Ródenas proposed an
online optimization model based on a discrete-event simulation model to provide
and support decisions about reassigning vehicles from other lines of the transport
system to the disturbed line. Meng et al. (2013) constructed a hybrid timed event
graph model for networked train operation simulation and timetable stability
optimization.
98 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

An assumption is made in the above publications, which is that the value range
boundaries of the fuzzy coefficients are identifiable. In the reality, especially in
engineering calculation, the value range boundaries of the fuzzy resources coeffi-
cients are not clear; sometimes they also have the fuzzy characteristics. In train
re-scheduling problem, the interval between the foregoing train’s departure from a
station and the backward train’s arrival can be seen as a fuzzy number, and even the
boundaries of the interval are fuzzy. Then it is necessary to study the train
re-scheduling problem with fuzzy linear programming model, in which the
right-hand side coefficients are fuzzy numbers, with the fuzzy value range
boundaries of the fuzzy coefficients.
In view of this fact, we will consider the problem under the fuzzy environment in
this chapter, which intends to make service strategies on the train re-scheduling
problem.
There are also numerous publications about the fuzzy linear programming
problem in recent years. The ANN was trained and tested with data extracted from
conflict resolutions in actual train operations in Turkish State Railways. A genetic
algorithm (GA) was developed to find the optimal solutions for small-sized prob-
lems in short times, and to reduce total delay times by around half in comparison to
the ANN. Fuzzy linear programming with fuzzy resource constraints coefficients is
a typical fuzzy linear programming. The key characteristic of this kind of pro-
gramming is that the coefficients of the resource are fuzzy, and the coefficients of
objectives are clear. The researchers have paid considerable attention to the
constraint-coefficient-linear fuzzy programming (Tanaka 1984; Delgado et al.
1993). Delgado et al. (1993) considered the use of nonlinear membership functions
in fuzzy linear programming problems to solve the linear programming problems
with fuzzy constraints. Gasimov and Yenilmez (2002) proposed the “modified
sub-gradient method” to solve linear programming problems with only fuzzy
technological coefficients and linear programming problems in which both the
right-hand side, and the technological coefficients were fuzzy numbers.
Ebrahimnejad (2011) generalized the concept of sensitivity analysis in fuzzy
number linear programming (FLNP) problems by applying fuzzy simplex algo-
rithms and using the general linear ranking functions on fuzzy numbers. Kazuo
(1984) proposed two extensions on the fuzzy linear programming proposed by
Zimmermann. He proved that fuzzy goals and fuzzy constraints expressed as fuzzy
relations with fuzzy parameters can be considered as fuzzy sets on different real
lines under some assumptions. And optimization in the case where the membership
functions of the fuzzy goals and the fuzzy constraints given in a piecewise linear
form can be achieved by using a standard linear programming technique. Frank
et al. (2008) proposed a fuzzy linear programming model which included opti-
mizing fuzzy constraints and objectives that consist of a triplet, and they gave a
modified simplex algorithm to address these problems.
Kaur and Kumar presented a new method to find the fuzzy optimal solution of
fully fuzzy path, i.e., critical path problems in which all the parameters are repre-
sented by LR flat fuzzy numbers. Dubey and Mehra (2014) proposed an approach
to model fuzzy multi-objective linear programming problems (FMOLPPs) from a
7.1 Introduction on Train Re-scheduling 99

perspective of bipolar view in preference modeling. Bipolarity was used to dis-

tinguish between the negative and the positive preferences.
Wan and Dong (2014) constructed an auxiliary multi-objective programming to
solve the corresponding possibility linear programming with trapezoidal fuzzy
numbers. Simic proposed a fuzzy risk explicit interval linear programming model
for end-of-life vehicle (ELV) recycling planning in the European Union, which had
advantages in reflecting uncertainties presented in terms of intervals in the ELV
recycling systems and fuzziness in decision-makers’ preferences. Ebrahimnejada
and Tavana (2014) proposed a new method for solving fuzzy linear programming
problems in which the coefficients of the objective function and the values of the
right-hand side are represented by symmetric trapezoidal fuzzy numbers while the
elements of the coefficient matrix are represented by real numbers. Rena and
Wangga (2014) considered a kind of bi-level linear programming problem where
the coefficients of both objective functions are fuzzy random variables and devel-
oped a computational method for obtaining optimistic Stackelberg solutions to such
a problem. Jin et al. (2014) developed a new Robust Inexact Joint-optimal a cut
Interval Type-2 Fuzzy Boundary Linear Programming (RIJ-IT2FBLP) model for
planning of energy systems by integrating both the interval T2 fuzzy sets and the
Inexact Linear Programming (ILP) methods. Kumar and Kaur (2013) pointed out
that the existing general form of such fully fuzzy linear programming problems in
which all the parameters are represented by such flat fuzzy numbers which are valid
only if there is no negative sign. They proposed a new general form of linear
programming to solve this problem. Kaur and Kumar (2013) also proposed fully
fuzzy linear programming (FLP) problems in which some or all the parameters are
represented by unrestricted L-R flat fuzzy numbers. Yano and Matsui (2013) pro-
posed an interactive decision-making method for hierarchical multi-objective fuzzy
random linear programming problems (HMOFRLP), in which multiple decision
makers in a hierarchical organization had their own multiple objective linear
functions with fuzzy random variable coefficients. Hajiagha et al. (2013). proposed
a model to extend the methodology for solving multi-objective linear programming
(MOLP) problems, when the objective functions and constraints coefficients are
stated as interval numbers. Fan et al. (2013) developed a generalized fuzzy linear
programming (GFLP) method for dealing with uncertainties expressed as fuzzy
sets. Sakawa and Matsui (2013) proposed an a-level sets of fuzzy random variables
and defined an a-stochastic two-level linear programming problem for guaranteeing
the degree of realization of the interactive fuzzy random cooperative two-level
linear programming problem. Dubey et al. (2012) studied the linear programming
problems involving interval uncertainty modeled using IFS. The nonmembership of
IFS was constructed with three different viewpoints viz., optimistic, pessimistic,
and mixed. Ebrahimnejad (2011) generalized the concept of sensitivity analysis of
fuzzy number linear programming (FLNP) problems by applying fuzzy simplex
algorithms and using the general linear ranking functions on fuzzy numbers.
These publications on fuzzy linear programming problems give us much
enlightenment on the application of the fuzzy linear programming in the
100 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

engineering computation. We improve the fuzzy linear programming in this chapter

and apply it in the train re-scheduling problem.
This chapter is structured as follows. Section 7.2 first introduces the problem in
a mathematical way, and then constructs a mathematical model under the specified
environment. In Sect. 7.3, to model the problem under the random fuzzy envi-
ronment, we improve the typical fuzzy linear programming with fuzzy resources.
Section 7.4 constructs the improved fuzzy linear programming model for train
re-scheduling. In Sect. 7.5, a computation case is presented, based on the data of
Beijing–Zhengzhou railway section to show the effectiveness of the model.
Section 7.6 draws conclusions.

7.2 Problem Statement

There are numerous methods for re-scheduling, including reduction of dwell time
on stations, reduction of the running time in sections, and change of the surpassing
stations. The goal is to recover the state in which the trains run according to the
planned timetable. In reality, the interval time and the buffer time are determined by
the train operating matrices. The elements of the inbound and outbound time matrix
are adjusted to change the running time in the sections, the dwell time at the stations
and the operation type when disruptions occur in real-world operations. This is the
essence of re-scheduling. A real-time re-scheduling plan must be proposed in a very
short period. On some occasions, the train’s track can coincide with the lines on the
planned timetable after some adjustments; sometimes it cannot.

7.2.1 Objective of Train Re-Scheduling Model

The goal of train operation adjustment is to make the actual dwell time accord with
the time as planned, when the trains are perturbed and delays occur. It is possible to
adjust the dwell of the trains so that there is a gap between the planned dwell and
the minimum time, as well as between the planned running time and the minimum
running time. The wider the gap, the less complicated the operating adjustment
work will be.
Thus, the train operation adjustment model with minimal summary delay time as
the destination can be defined as follows:

X
N X
M
min z ¼ ½maxðai;k  a0i;k ; 0Þ þ ðdi;k  di;k
0
Þ; ð7:1Þ
i¼1 k¼1

where ai;k stand for the inbound time of train i at station k and di;k stand for the
outbound time of train i at station k. a0i;k and di;k
0
stand for the original planned
7.2 Problem Statement 101

inbound and outbound times respectively. Then the objective is to minimize the gap
between the re-scheduled timetable and the original timetable. Because that a
passenger train can arrive at a station earlier than it is planned and we do not care
about it when calculating the delayed time, the gap between the re-scheduled arrival
time and the original planned arrival time is described as maxðai;k  a0i;k ; 0Þ. And
di;k  di;k
0
is a constraint for the passenger trains, so the gap between the
re-scheduled departure time and the original departure time is set to be di;k  di;k
0
.

7.2.2 System Constraints

There are numerous prerequisite rules in railway operation to ensure the safety,
which are determined by the facilities such as the blocking systems. The most
important rule is to determine the relations between the inbound and outbound time
of all the trains, to separate the trains in space. So the system constraints are
designed as follows.
The difference between a backward train arriving time and a forward train
arriving time at the same stations must be longer than the technical intervals, which
produces the constraint.

ai þ 1;k  ai;k  Ia ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M; ð7:2Þ

where Ia is the interval time between the two inbound times of train i and train i + 1
at Station k.
Likewise, the difference between a backward train departing time and a forward
train departing time from the same stations must be longer than the technical
intervals. The constraint can be described as

di þ 1;k  di;k  Id ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M; ð7:3Þ

where Id is the interval time between the two outbound times of train i and train
i + 1 at Station k.
The interval between two trains must satisfy the departing arriving interval and
arriving departing interval. Set sdepartarrive to be the minimum time interval between
a train leaving a station and another train arrival the same station. The constraints
are defined in Eq. (7.4).

ai þ 1;k  di;k [ sdepartarrive ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M ð7:4Þ

The running time of each train according to the re-scheduled timetable must be
longer than the minimum running time, which can be formulated as follows:
102 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

min;run
ai;k þ 1  di;k  ti;k ; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M  1; ð7:5Þ

min;run
where ti;k is the minimum time of train i on the section between station k and
k + 1.
Again, the dwelling time of each train must be longer than the minimum
dwelling time, which produces the constraint
min;dwell
di;k  ai;k  ti;k ; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M; ð7:6Þ

min;dwell
where ti;k is the minimum dwelling time of train i at station k.
Passenger trains must not leave the stations before the time as it is planned on the
timetable, which is made available to the public. So there is a constraint as follows:

di;k  di;k
0
 0; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M ð7:7Þ

7.2.3 Mathematical Model

The mathematical model of this problem is constructed as follows:

8
> PN P M
>
> minz ¼ max½ðai;k  a0i;k ; 0Þ þ ðdi;k  di;k
0
Þ
>
>
>
> i¼1 k¼1
> s:t:
>
>
>
>
< ai þ 1;k  ai;k  Ia ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
>
di þ 1;k  di;k  Id ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M ð7:8Þ
>
> a i þ 1;k  di;k [ sdepartarrive ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
>
>
>
>   min;run
; ¼ 2; . . .; N; k ¼ 1; 2; . . .; M  1
>
> a i;k þ 1 d i;k t i;k i 1;
>
>
>
> d  ai;k  ti;k min;dwell
; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M
>
: i;k
di;k  di;k0
 0; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M

where ai;k and di;k are the inbound and outbound times of the ith train at kth station.
It is easy to see that all the parameters in the model (8) are supposed to be fixed
quantities. In the authentic conditions, a rescheduled timetable is usually designed
after the occurrence of an emergency. Thus, the concrete values of some parameters
actually cannot be obtained in advance, especially the parameters on the right side
of the constraints equations. To deal with the problem in a mathematical way, we
usually treat these parameters as fuzzy variables according to the experts’ experi-
ence when we cannot get enough real sample data to calculate out the parameters by
statistical ways.
To solve the model, we changed the styles of the objective and the constraints
into standard styles, reconstructing the model (7.8) as follows:
7.2 Problem Statement 103

8
> P
N P M
>
> maxz ¼ Cmax  ½maxðai;k  a0i;k ; 0Þ þ ðdi;k  di;k
0
Þ
>
>
>
> i¼1 k¼1
> s:t:
>
>
>
>
< ai;k  ai þ 1;k   Ia ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
>
di;k  di þ 1;k   Id ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M ð7:9Þ
>
> d i;k  ai þ 1;k \sdepartarrive ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
>
>
>
>    min;run
; ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M  1
>
> d i;k a i;k þ 1 t i;k i
>
>
> ai;k  di;k   ti;k
>
min;dwell
; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M
>
: 0
di;k  di;k  0; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M

7.3 Fuzzy Linear Programming with Fuzzy Resource

Constraints Theory

7.3.1 Fuzzy Linear Programming with Fuzzy Resources

Typical fuzzy linear programming with fuzzy resources can be described as

follows:
8
> max z ¼ cT x
>
< s:t:
Ax  b ð7:10Þ
>
>
: 
x0

where b is the fuzzy resources coefficients vector. For the fuzzy constraints Ax  b,

the ith constraint is

ðAxÞi  bi ; i ¼ 1; 2; . . .; m ð7:11Þ


Set the maximal tolerance to be pi , the fuzzy function is defined as follows:

8
< 1; ðAxÞi  bi
a ¼ li ðxÞ ¼ 0; ðAxÞi [ bi þ pi ð7:12Þ
:
1  ððAxÞi  bi Þ=pi ; bi  ðAxÞi  bi þ pi

Then, the fuzzy linear programming model (1) can be remodeled as follows:
8
>
> max z ¼ cT x
<
s:t:
ð7:13Þ
>
> ðAxÞi  bi þ ð1  aÞpi ; i  1; 2; . . .; m
:
x0
104 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

Set h ¼ 1  a, the model is turned to be

8
>
> max z ¼ cT x
<
s:t:
ð7:14Þ
>
> Ax  b þ hp
:
x0

The optimal solution to the model is

~S ¼ fðx ðhÞ; 1  hÞjh 2 ½0; 1g ð7:15Þ

7.3.2 Fuzzy Linear Programming of Resources with Fuzzy

Coefficients Boundaries of Fuzzy Resources

There is a kind of fuzzy linear programming problem for the engineering compu-
tation, which has the fuzzy boundaries of the coefficients value range. The
boundaries can be described as fuzzy numbers. That is to say, the upper and lower
boundaries of b are fuzzy numbers.
The ith constraint of Ax  b is


ðAxÞi  bi ; i ¼ 1; 2; . . .; m ð7:16Þ


Then, bi 2 ½Li ; Ui , Li ; Ui are the lower and upper boundaries of bi respectively.

Set

bi ¼ f ðUi Þ; ð7:17Þ

where f is a fuzzy membership function. It can be a triangle, a trapezoid or

Gaussian function.
Then, Ui can be described as

Ui ¼ f 1 ðbi Þ; ð7:18Þ

where f 1 is the inverse function of f .

Likewise, set

ci ¼ gðLi Þ ð7:19Þ

g is a kind of fuzzy membership function, similar with f .

7.3 Fuzzy Linear Programming with Fuzzy Resource Constraints Theory 105

Then
Li ¼ g1 ðci Þ; ð7:20Þ

where g1 is the inverse function of g.

As it is known,bi 2 ½Li ; Ui , then set

a ¼ lðbi Þ ¼ qðLi ; Ui Þ; ð7:21Þ

where q and l are similar with f . They are also fuzzy membership functions. Then
bi can be described as

bi ¼ q1 ðLi ; Ui Þ ¼ q1 ðg1 ðci Þ; f 1 ðbi ÞÞ; ð7:22Þ

where q1 is the inverse function of q.

So the model can be changed as
8
> max z ¼ cT x
>
< s:t:
ðAxÞi  q1 ðLi ; Ui Þ ¼ q1 ðg1 ðci Þ; f 1 ðbi ÞÞ; i ¼ 1; 2; . . .; m ð7:23Þ
>
>
: 
x0

7.4 Improved Fuzzy Linear Programming Model

for Train Re-scheduling

By using the theory in Sect. 7.3, model (7.9) can be remodeled as a fuzzy linear
programming model with fuzzy resources’ boundaries constraints.
The sdepartarrive is a fuzzy number. Even the boundaries of value range of
sdepartarrive are fuzzy. So we can remodel the problem as follows:
8 M h i
> PN P
>
> maxz ¼ Cmax  maxðai;k  a0i;k ; 0Þ  ðdi;k  di;k
0
Þ
>
>
>
> i¼1 k¼1
>
> s:t:
>
>
>
> ai;k  ai þ 1;k   Ia ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
<
di;k  di þ 1;k   Id ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
>
> di;k  ai þ 1;k \q1 ðg1 ðcÞ; f 1 ðbÞÞ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
>
>
>
> min;run
di;k  ai;k þ 1   ti;k ; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M  1
>
>
>
>
>
> a  d   t min;dwell
; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M
>
: 0
i;k i;k i;k
di;k  di;k  0; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M
ð7:24Þ
106 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

where c is the membership of the lower boundary of sdepartarrive and b is the

membership of upper boundary of sdepartarrive . g1 ðcÞ is the lower boundary of
sdepartarrive and f 1 ðbÞ is the upper boundary of sdepartarrive . Cmax is an enough
P
N P
M
k¼1 maxðai;k  ai;k ; 0Þ and di;k  di;k .
0 0
large number which is much larger than
i¼1
In this chapter, the membership functions, q, g and f are designed as the triangle
functions. The upper bound of sdepartarrive is U, which is a triangle fuzzy member.
U  ½2; 4, and the average value of U is 3. The lower bound of sdepartarrive is L,
whose average value is 2. b is the value of membership.
8
>
> 0; U  4
<
ð4  UÞ=ð4  3Þ; 3  U\4
b ¼ f ðUÞ ¼ ð7:25Þ
> ðU1  2Þ=ð3  2Þ; 2\U\3
>
:
0; U1  2

Since the constraints in this problem are resources constraints, then the more the
resources are, the bigger the objective value will be. So when the membership value
is b, the left part of the polyline in Fig. 7.1 is useless when solving the problem.
Then the useful part is kept, as below.

0; U  4
b ¼ f ðUÞ ¼ ð7:26Þ
ð4  UÞ=ð4  3Þ; 3  U\4

Then the upper bound of sdepartarrive is U ¼ 4  b.

In the like manner, c is set to be the membership value for the lower bound of
sdepartarrive .

0; L  3
c ¼ gðLÞ ¼ ð7:27Þ
ð3  LÞ=ð3  2Þ; 2  L\3

The lower bound of sdepartarrive is L ¼ 3  c.

Additionally, sdepartarrive is also described as a triangle fuzzy number. The
membership value a is

Fig. 7.1 The left line is not β

available when the objective
function is triangle function 1

0 2 3 4 U
7.4 Improved Fuzzy Linear Programming Model for Train Re-scheduling 107

(
0; sdepartarrive  U
a ¼ lðsdepartarrive Þ ¼ 
ðU  sdepartarrive Þ ðU  ðU þ LÞ=2Þ; ðU þ LÞ=2  sdepartarrive \U
ð7:28Þ

Then we get
1
sdepartarrive ¼ 4  b  að1  b þ cÞ
2

Then the model can be changed into

8
> PN P M
>
> maxz ¼ Cmax  ½maxðai;k  a0i;k ; 0Þ  ðdi;k  di;k
0
Þ
>
>
>
> i¼1 k¼1
>
> s:t:
>
>
>
> i;k  ai þ 1;k   Ia ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
a
<
di;k  di þ 1;k   Id ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
>
> i;k  ai þ 1;k \4  b  12 að1  b þ cÞ; i ¼ 1; 2; . . .; N  1; k ¼ 1; 2; . . .; M
d
>
>
>
> di;k  ai;k þ 1   ti;kmin;run
; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M  1
>
>
>
>
>
> a  d   t min;dwell
; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M
>
: i;k i;k i;k
di;k  di;k  0; i ¼ 1; 2; . . .; N; k ¼ 1; 2; . . .; M
0

ð7:29Þ

In this model, a resource constraint is analyzed and seen as a fuzzy constraint.

Even the boundaries of the resource sdepartarrive are described as the fuzzy numbers.
All the fuzzy numbers are transferred into certain numbers to be ready to be solved.
This treatment makes the constraint and the model more accordance with the actual
case. In reality, some of the resources are actually very difficult to obtain and have
the fuzzy characteristics. So it is necessary to do such process. The most important
improvement compared to the existing fuzzy linear model for train re-scheduling is
also such process.

7.5 Computation Cases and Analysis

7.5.1 Results of the Computation Cases

There are 13 stations and 12 sections on the section between Beijing and
Zhengzhou. We apply the model into two computation cases. The first one is that
we assume several trains are delayed for a period of time. The second one is that we
assume that a track in a section is affected by an emergency and the two-track
railway section becomes a single-track railway.
108 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

Fourteen trains are planned at the down-going direction and another 14 trains at
the up-going direction on the working diagram from 8 a.m. to 12. a.m. The original
planned timetable is shown in Tables 7.1 and 7.2, and Fig. 7.2. The minimum
dwelling time of all the trains at each station are listed in Table 7.3. The data in
Table 7.4 are the minimum running time of all the trains in each railway section.
The trains are divided into two grades. G71, G83, G79, G90, G92 belong to the first
grade, which requires less running time on each section that the trains G507, G651,
G501, 653, G509, G571, G511, G655, G513, G657, G515, G560, G508, G562,
G652, G502, G654, G512, G672, G6732, G6734, G6704, G602, and G92 which
belong to the second grade.
Case 1 In computation Case 1, we take it for granted that five trains at the
down-going direction, G83, G571, G511, G79, G655, and four trains at the
up-going direction, G90, G508, G562, G652 are disturbed when running on section
between Beijing and Zhuozhou. They are later 9, 13.5, 10, 10.5, 20, 10, 10, 10, and
10 min respectively than as planned.
In the computation, we use a computer with a CPU of i5-2400 and 2G RAM.
The software is Matlab 6.0.
In this experiment, the optimal solution is obtained with the parameters a = 1,
min;run min;dwell
b ¼ 1. Ia ¼ Id ¼ 3. ti;k and ti;k are set as the data shown in Tables 7.3 and
7.4.
Since Cmax is set to be 5000, the optimal objective of the model is calculated to
be 4197.5 according to model (7.29). The summary delayed time of the down-going
trains is 485 min and the summary delayed time of up-going trains is 317.5 min.
The total delay time of all the trains at all the stations is 802.5 min, including the
arriving delay time and the departing delay time.
For different groups of parameters, the computational results are presented in
Fig. 7.3.
According to the parameter linear programming algorithm, the solution to the
train re-scheduling model is shown in Tables 7.5 and 7.6, and Fig. 7.4. The
inbound and outbound times in Tables 7.5 and 7.6 in italic type are the re-scheduled
time based on the data in Tables 7.1 and 7.2.
It is easy to see that all the delayed trains recover the operation according to the
original timetable before 11:40. G83 is planned to overtake G509 at Dingzhou at
9.5830. Since G83 arrived late at Shijiazhuang station, it is designed to overtake
G509 at Shijiazhuang according to replanned timetable. It recovers to operate
according to the original timetable at Hebi at 11:0330. The other four trains at the
down-going direction eliminate the delays at Shijiazhuang station.
G90 is re-scheduled to reduce the delayed time in the whole section and arrives
at Beijing in time as it is planned. It still dwells on Shijiazhuang for 2 min. G560 is
affected by G90 because the minimum interval between departures is 3 min and
G90 departures from Shijiazhuang at 9:5800. G560 has to start off not earlier than
10:01. G508 and G562 recover the operation according to the original timetable at
Shijiazhuang station and G652 fulfills the process at Gaoyi station.
Table 7.1 The planned timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the down-going direction
G507 G651 G501 G71 G653
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 8.0000 8.2100
Zhuozhou 8.2400 8.2400 8.4600 8.5400
Gaobeidian 8.2900 8.2900 8.5900 8.5900
Baoding 7.4600 7.4800 8.0900 8.0900 8.4100 8.4300 9.1100 9.1100
Dingzhou 7.5400 7.5600 8.0800 8.0800 8.2400 8.2600 8.5800 8.5800 9.2600 9.2600
Shijiazhuang 8.1900 8.2300 8.2400 8.2800 8.4900 8.5200 9.1900 9.2200 9.4600 9.4900
Gaoyi 8.3700 8.3700 8.4200 8.4200 9.0600 9.0600 9.3600 9.3600 10.0100 10.0100
7.5 Computation Cases and Analysis

Xingtai 8.5200 8.5200 8.5600 8.5800 9.2000 9.2200 9.5100 9.5100 10.1500 10.1500
Handan 9.0200 9.0400 9.1400 9.1600 9.3200 9.3200 10.0100 10.0300 10.2400 10.2400
Anyang 9.1800 9.1800 9.3000 9.3000 9.4700 9.4900 10.1700 10.1700 10.3700 10.3900
Hebi 9.3100 9.3300 9.4400 9.4600 10.0100 10.0100 10.3000 10.3000 10.5300 10.5300
Xinxiang 9.4900 9.5200 9.5700 9.5700 10.1100 10.1100 10.4100 10.4300 11.0500 11.1800
Zhengzhou 10.1300 10.1900 10.3100 11.0400 11.3900
G509 G83 G571 G511 G79
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 8.4300 9.0000 9.2700 9.3700 10.0000
Zhuozhou 9.0700 9.0700 9.2100 9.2100 9.4630 9.4630 10.0200 10.0400 10.2130 10.2130
Gaobeidian 10.1100 10.1100 9.2500 9.2500 9.5200 9.5200 10.0830 10.0830 10.2530 10.2530
Baoding 9.2400 9.2600 9.3600 9.3600 10.0800 10.1000 10.2100 10.2100 10.3700 10.3700
Dingzhou 9.4400 9.5100 9.4830 9.4830 10.2500 10.2500 10.3600 10.3600 10.4950 10.4950
Shijiazhuang 10.1400 10.1700 10.0700 10.0900 10.4600 10.5000 10.5600 10.5900 11.0700 11.0900
Gaoyi 10.3000 10.3000 10.2030 10.2030 11.0330 11.0330 11.1300 11.2500 11.2030 11.2030
Xingtai 10.4300 10.4300 10.3200 10.3200 11.1800 11.2000 11.4230 11.4230 11.3200 11.3200
(continued)
109
Table 7.1 (continued)
110

Handan 10.5400 10.5400 10.4100 10.4100 11.3600 11.4500 11.5500 11.5500 11.4030 11.4030
Anyang 11.0600 11.0600 10.5230 10.5230 11.5800 11.5800 11.5200 11.5200
Hebi 11.1800 11.2000 11.0330 11.0330
Xinxiang 11.3000 11.3000 11.1200 11.1200
Zhengzhou 11.5000 11.3000
G655 G513 G657 G515
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
Beijing 10.0500 10.4800 11.0600 11.5000
7

Zhuozhou 10.3000 10.3000 11.1300 11.1300 11.3000 11.3000

Gaobeidian 10.3500 10.3700 11.1800 11.2000 11.3500 11.3500
Baoding 10.5300 10.5500 11.3130 11.3130 11.4700 11.4900
Dingzhou 11.1000 11.1000 11.4500 11.4500
Shijiazhuang 11.3100 11.3400 12.0700 12.1100
Gaoyi 11.5030 11.5030
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
Train Re-scheduling Based on an Improved Fuzzy Linear …
Table 7.2 The planned timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the up-going direction
G560 G90 G508 G562 G652
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 11.3000
Zhuozhou 11.3700 11.3900 11.0900 11.0900
Gaobeidian 11.2600 11.2800 11.0430 11.0430
Baoding 11.0700 11.1000 10.5300 10.5300 11.5100 11.5100
Dingzhou 10.5300 10.5300 10.4100 10.4100 11.3700 11.3700 11.5130 11.5130
Shijiazhuang 10.3400 10.3400 10.2100 10.2300 11.1300 11.1700 11.2800 11.3100 11.5500 11.5800
Gaoyi 10.2030 10.2030 10.0830 10.0830 11.0000 11.0000 11.1400 11.1400 11.3800 11.4000
7.5 Computation Cases and Analysis

Xingtai 10.0700 10.0700 9.5600 9.5600 10.4300 10.4500 10.5930 10.5930 11.2400 11.2400
Handan 9.5500 9.5700 9.4700 9.4700 10.3300 10.3300 10.4700 10.4900 11.1400 11.1400
Anyang 9.3000 9.3800 9.3400 9.3400 10.1830 10.1830 10.2800 10.3000 11.0000 11.0000
Hebi 9.1300 9.1500 9.2400 9.2400 10.0300 10.0500 10.1300 10.1300 10.4500 10.4700
Xinxiang 8.5500 8.5700 9.1600 9.1600 9.5330 9.5330 10.0000 10.0200 10.3400 10.3400
Zhengzhou 8.3500 9.0000 9.3200 9.4000 10.1400
G502 G654 G512 G672 G6732
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 10.2300
Zhuozhou 9.5900 9.5900
Gaobeidian 9.5430 9.5430
Baoding 9.3900 9.4200
Dingzhou 9.2400 9.2400
Shijiazhuang 9.0000 9.0300
Gaoyi 8.4300 8.4500
Xingtai 8.1730 8.1730
(continued)
111
Table 7.2 (continued)
112

Handan 12.0000 12.0000 8.0600

Anyang 11.5600 11.5600 11.4800 11.4800
Hebi 11.4400 11.4400 11.3830 11.3830
Xinxiang 11.2100 11.3400 11.2900 11.2900 11.5000 11.5200 12.0000 12.0200
Zhengzhou 11.0100 11.1200 11.3000 11.4000
G6734 G6704 G602 G92
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
Beijing 12.1900 8.5700 9.5300 11.0000
7

Zhuozhou 11.5200 11.5400 8.2900 8.3200 9.2900 9.2900 10.3800 10.3800

Gaobeidian 11.4100 11.4300 8.1800 8.2000 9.2500 9.2500 10.3400 10.3400
Baoding 11.2800 11.2800 8.0100 9.1000 9.1200 10.2100 10.2100
Dingzhou 11.1330 11.1330 8.5600 8.5600 10.0900 10.0900
Shijiazhuang 10.5100 10.5300 8.3400 9.5000
Gaoyi 10.3730 10.3730
Xingtai 10.2300 10.2300
Handan 10.1300
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
Train Re-scheduling Based on an Improved Fuzzy Linear …
7.5 Computation Cases and Analysis 113

Stations G655
G71 G653 G509 G83 G571 G511 G79 G513 G657 G515
Beijing 1 3 3 0 6
70 7 3 0 5 8 0 0
G501
2 6 9 92 8 9 9 4
Zhuozhou 0 45
Gaobeidian 0 4 9 4 7 15 6.30 4.30 4 1.30 4.30 3 8 80 7 3 2
G651 8 9 9 1 5 2 8.30 5.30 7 0 6 5 1
Baoding 1 1 2 4 2 8 1 33 0 8 71
G507 9 3 01 6 6 9 0 1 7 5 7 1.30 9
4 4 4 9 1 3 3.30 7 1.30
Dingzhou 8 6 68 6 8.30 1 5 6 9.30 0 5
G6704
9 4 4 9 3 9 60 7 4 3 4 6 3 6 7 7 11 8
Shijiazhuang G92
3 8 2 0 2 9 9 7 1 10 9 9 3 8 4 5
G602 34
5 8.30 0.30 7.30 0 0
Gaoyi 37 2 6 6 1 0.30 0 0.30 5 8 0.30

Xingtai 7.30 6 0 6 7 3 5 9.30 8 4

2 8 2 1 5 2 33 0 2 2.30
Handan 6 2 4 7 7 3 3 9 4 6
4 6 2 5 3 4 1 7 4 0.30 5 5 0
G6732 G6734 8.30
Anyang 8 7 0 7 0 8 6
8 00 4 9 7 8 9 2.30 6 2 8
Hebi 5 4 1 4 5 3 7 8.30 4
3 3 6 1 3 0 5 3 3.30 0
7 6 9 3.30 2 4 1 5 94 2
Xinxiang 5 2 70 1 3 2 8 1 0 0 0

Zhengzhou 5 2 0 34 9 1 1 4 2 0 90 0
8:00 10:00 11:00 G654 G512 12:00
G560 9:00 G508
G90 G562 G652 G502 G672 Time

Fig. 7.2 The planned train working diagram from 8 to 12 a.m. in section between Beijing and
Zhengzhou

In addition, it is needed to analyze the sensitivity of optimal objectives with

respect to the parameters. We set a and b respectively and solve the fuzzy linear
programming, as shown in Fig. 7.3. We can see the larger the membership is, the
larger the objective value is. When a ¼ b ¼ 1, the programming degenerates into a
typical fuzzy linear programming. With a and b rise, the objective value becomes
larger and larger. When a ¼ b ¼ 1, the objective value is the maximal, 4197.5. At
this point, the membership is the biggest. The interpretation is, the objective value is
reaching the maximal value gradually with the more possibility that the objective
coefficient is set to be the biggest.
To show the advancement of the improved fuzzy linear programming proposed
in this chapter, we also did the data experiments with the typical fuzzy linear
programming model. The re-scheduled timetables are shown in Tables 7.7 and 7.8
and Fig. 7.5.
Case 2 In this case, the relevant basic data are the same as those in computation.
We assume that an emergency occurs in the section between Beijing and Zhuozhou,
causing a failure of one of the tracks. Then the section between Beijing and
Zhuozhou becomes a single-track rail section. The time interval for two meeting
trains at a station is set to be 1 min. Then the computing results of Case 2 are listed
in Tables 7.9 and 7.10 and Fig. 7.6.
According to the computing results, the departure from Zhuozhou of G6704 is
delayed for 15 min to avoid the conflict with G563. Then the chain reaction is
caused for the trains from different directions cannot occupy the affected section
simultaneously. G509, G83 are postponed for 23.5 and 12 min respectively to
 min;dwell
Table 7.3 The minimal dwelling time of all the trains at each station (ti;k )
114

G507 G651 G501 G71 G653 G509 G83 G571 G511 G79 G655 G513 G657 G515
Beijing – – – – – – – – – – – – – –
Zhuozhou – – – 0 8 0 0 0 2 0 0 0 0 –
Gaobeidian – – – 0 0 0 0 0 0 0 2 2 0 –
Baoding – 2 0 2 0 2 0 2 0 0 0 0 2 –
Dingzhou 2 0 2 0 0 7 0 0 0 0 0 0 – –
Shijiazhuang 4 4 3 3 3 3 2 4 3 2 4 4 – –
Gaoyi 0 0 0 0 0 0 0 0 12 0 – – – –
7

Xingtai 0 2 2 0 0 0 0 2 0 0 – – – –
Handan 2 2 0 2 0 0 0 9 0 0 – – – –
Anyang 0 0 2 0 2 0 0 0 – 0 – – – –
Hebi 2 2 0 0 0 2 0 – – – – – – –
Xinxiang 3 0 0 2 13 0 0 – – – – – – –
Zhengzhou – – – – – – – – – – – – – –
G560 G90 G508 G562 G652 G502 G654 G512 G672 G6732 G6734 G6704 G602 G92
Beijing – – – – – – – – – – – – – –
Zhuozhou 0 0 – – – – – – – 0 2 3 0 0
Gaobeidian 0 0 – – – – – – – 0 2 2 0 0
Baoding 3 0 0 – – – – – – 3 0 – 2 0
Dingzhou 0 0 0 0 – – – – – 0 0 – 0 0
Shijiazhuang 0 2 2 3 3 – – – – 3 2 – – 0
Gaoyi 0 0 0 0 2 – – – – 4 0 – – –
Xingtai 0 0 2 0 0 – – – – 0 0 – – –
Handan 2 0 0 2 0 – – – – – 0 – – –
Anyang 2 0 0 2 0 – 0 – – – – – – –
Hebi 2 0 2 2 2 0 0 – 0 – – – – –
Xinxiang 2 0 0 2 0 2 0 0 – – – – – –
Zhengzhou – – – – – – – – – – – – – –
Train Re-scheduling Based on an Improved Fuzzy Linear …
7.5 Computation Cases and Analysis 115

min;run
Table 7.4 The minimal running time of all the trains in each section (ti;k )
G507, G651, G501, G653, G509, G71, G83,
G571, G511, G655, G513, G657, G79, G90, G92
G515, G560, G508
G562, G652, G502, G654, G512,
G672, G6732, G6734, G6704, G602, G92
Beijing-Zhuozhou 21 20
Zhuozhou-Gaobeidian 4 3
Gaobeidian-Baoding Baoding 9 8
Baoding- Dingzhou 10.30 9
Dingzhou- Shijiazhuang 12.30 11
Shijiazhuang-Gaoyi 10 9.30
Gaoyi-Xingtai 12 10.30
Xingtai-Handan 9 7
Handan-Anyang 12 10
Anyang-Hebi 12 9.30
Hebi-Xinxiang 10 8.30
Xinxiang-Zhengzhou 20 18
note dd.ee stands for dd minutes and ee seconds

Fig. 7.3 Relation between Relation between objective function value and α,β
objective function value and
a, b
5000

4000

3000

2000

1000

0
1
0.9
0.8
0.7
0.6 0.9 1
0.5 0.7 0.8
0.4 0.6
0.3 0.4 0.5
0.2 0.3
0.1
0 0 0.1 0.2
β α

avoid G6704, and G79, G655 are delayed to avoid G6732, G92 is delayed at
Zhuozhou to avoid the conflict with G79 and G655 and so on. G511 is re-scheduled
to arrive at Zhuozhou earlier than it is planned to assure that G6732 can run as it is
planned.
The summary delayed time of the down-going trains is 1625 min and that of the
up-going trains is 125 min. It is because that the emergency occurs at the section
Table 7.5 The re-scheduled timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the down-going direction in computation Case 1
116

G507 G651 G501 G71 G653

Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 8.0000 8.2100
Zhuozhou 8.2400 8.2400 8.4600 8.5400
Gaobeidian 8.2900 8.2900 8.5900 8.5900
Baoding 7.4600 7.4800 8.0900 8.0900 8.4100 8.4300 9.1100 9.1100
Dingzhou 7.5400 7.5600 8.0800 8.0800 8.2400 8.2600 8.5800 8.5800 9.2600 9.2600
Shijiazhuang 8.1900 8.2300 8.2400 8.2800 8.4900 8.5200 9.1900 9.2200 9.4600 9.4900
7

Gaoyi 8.3700 8.3700 8.4200 8.4200 9.0600 9.0600 9.3600 9.3600 10.0100 10.0100
Xingtai 8.5200 8.5200 8.5600 8.5800 9.2000 9.2200 9.5100 9.5100 10.1500 10.1500
Handan 9.0200 9.0400 9.1400 9.1600 9.3200 9.3200 10.0100 10.0300 10.2400 10.2400
Anyang 9.1800 9.1800 9.3000 9.3000 9.4700 9.4900 10.1700 10.1700 10.3700 10.3900
Hebi 9.3100 9.3300 9.4400 9.4600 10.0100 10.0100 10.3000 10.3000 10.5300 10.5300
Xinxiang 9.4900 9.5200 9.5700 9.5700 10.1100 10.1100 10.4100 10.4300 11.0500 11.1800
Zhengzhou 10.1300 10.1900 10.3100 11.0400 11.3900
G509 G83 G571 G511 G79
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 8.4300 9.0000 9.2700 9.3700 10.0000
Zhuozhou 9.0700 9.0700 9.3000 9.3000 10.0000 10.0000 10.1200 10.1400 10.3200 10.3200
Gaobeidian 10.1100 10.1100 9.3530 9.3530 10.0400 10.0400 10.1600 10.1600 10.3500 10.3500
Baoding 9.2400 9.2600 9.4400 9.4400 10.1600 10.1800 10.2700 10.2700 10.4330 10.4330
Dingzhou 9.4400 9.5100 9.5800 9.5800 10.2930 10.2930 10.3900 10.3900 10.5330 10.5330
Shijiazhuang 10.1400 10.2200 10.1700 10.1900 10.4600 10.5000 10.5600 10.5900 11.0700 11.0900
Gaoyi 10.3200 10.3200 10.2900 10.2900 11.0330 11.0330 11.1300 11.2500 11.2030 11.2030
Xingtai 10.4500 10.4500 10.3930 10.3930 11.1800 11.2000 11.4230 11.4230 11.3200 11.3200
(continued)
Train Re-scheduling Based on an Improved Fuzzy Linear …
Table 7.5 (continued)
Handan 10.5400 10.5400 10.4700 10.4700 11.3600 11.4500 11.5500 11.5500 11.4030 11.4030
Anyang 11.0600 11.0600 10.5700 10.5700 11.5800 11.5800 11.5200 11.5200
Hebi 11.1800 11.2000 11.0700 11.0700
Xinxiang 11.3000 11.3000 11.1430 11.1430
Zhengzhou 11.5000 11.3000
G655 G513 G657 G515
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
Beijing 10.0500 10.4800 11.0600 11.5000
Zhuozhou 10.5000 10.5000 11.1300 11.1300 11.3000 11.3000
7.5 Computation Cases and Analysis

Gaobeidian 10.5500 10.5700 11.1800 11.2000 11.3500 11.3500

Baoding 11.0600 11.0800 11.3130 11.3130 11.4700 11.4900
Dingzhou 11.1830 11.1830 11.4500 11.4500
Shijiazhuang 11.3100 11.3400 12.0700 12.1100
Gaoyi 11.5030 11.5030
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
The time in italics is the rescheduled inbound and outbound time of the trains at stations
117
Table 7.6 The re-scheduled timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the up-going direction in computation Case 1
118

G560 G90 G508 G562 G652

Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 12.0400 11.3000
Zhuozhou 11.3700 11.3900 11.1200 11.1200
Gaobeidian 11.2600 11.2800 11.0830 11.0830
Baoding 11.0700 11.1000 10.5900 10.5900 11.5100 11.5100 12.0600 12.0800
Dingzhou 10.5400 10.5400 10.4800 10.4800 11.3700 11.3700 11.5130 11.5130
Shijiazhuang 10.3530 10.3530 10.3100 10.3300 11.1300 11.1700 11.2800 11.3100 11.5500 11.5800
7

Gaoyi 10.2300 10.2300 10.1900 10.1900 11.0030 11.0030 11.1400 11.1400 11.3800 11.4000
Xingtai 10.1000 10.1000 10.0700 10.0700 10.4500 10.4700 10.5930 10.5930 11.2630 11.2630
Handan 9.5500 10.0100 9.5800 9.5800 10.3730 10.3730 10.4800 10.5000 11.1830 11.1830
Anyang 9.3000 9.3800 9.4600 9.4600 10.2500 10.2500 10.3100 10.3300 11.0730 11.0730
Hebi 9.1300 9.1500 9.3500 9.3500 10.1200 10.1400 10.2030 10.2030 10.5500 10.5700
Xinxiang 8.5500 8.5700 9.2600 9.2600 10.0330 10.0330 10.1000 10.1200 10.4400 10.4400
Zhengzhou 8.3500 9.0000 9.3200 9.4000 10.1400
G502 G654 G512 G672 G6732
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 10.2300
Zhuozhou 9.5900 9.5900
Gaobeidian 9.5430 9.5430
Baoding 9.3900 9.4200
Dingzhou 9.2400 9.2400
Shijiazhuang 9.0000 9.0300
Gaoyi 8.4300 8.4500
Xingtai 8.1730 8.1730
(continued)
Train Re-scheduling Based on an Improved Fuzzy Linear …
Table 7.6 (continued)
Handan 12.0000 12.0000 8.0600
Anyang 11.5600 11.5600 11.4800 11.4800
Hebi 11.4400 11.4400 11.3830 11.3830
Xinxiang 11.2100 11.3400 11.2900 11.2900 11.5000 11.5200 12.0000 12.0200
Zhengzhou 11.0100 11.1200 11.3000 11.4000
G6734 G6704 G602 G92
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
Beijing 12.1900 8.5700 9.5300 11.0000
Zhuozhou 11.5200 11.5400 8.2900 8.3200 9.2900 9.2900 10.3800 10.3800
7.5 Computation Cases and Analysis

Gaobeidian 11.4100 11.4300 8.1800 8.2000 9.2500 9.2500 10.3400 10.3400

Baoding 11.2800 11.2800 8.0100 9.1000 9.1200 10.2100 10.2100
Dingzhou 11.1330 11.1330 8.5600 8.5600 10.0900 10.0900
Shijiazhuang 10.5100 10.5300 8.3400 9.5000
Gaoyi 10.3730 10.3730
Xingtai 10.2300 10.2300
Handan 10.1300
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
The time in italics is the rescheduled inbound and outbound time of the trains at stations
119
120 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

Stations G655
G71 G653 G509 G83 G571 G511 G79 G513 G657 G515
Beijing

G501

Zhuozhou
Gaobeidian
G651
Baoding
G507

Dingzhou
G6704
Shijiazhuang 3
G92 1
G602
Gaoyi

Xingtai
Handan 8 1
G6732 G6734
Anyang

Hebi

Xinxiang

Zhengzhou
8:00 9:00 10:00 11:00 G654 G512 12:00
G560 G90 G508 G672 Time
G562 G652 G502

Fig. 7.4 The replanned train working diagram generated with the improved fuzzy linear
programming model from 8 to 12 a.m. in section between Beijing and Zhengzhou in computation
Case 1. The dotted lines stand for the original planned moving trajectories of the disturbed trains.
The red lines are the re-scheduled moving trajectories of the disturbed trains. The wavy lines imply
that the trains are disturbed when running in section between Beijing and Zhuozhou

between Beijing and Zhuozhou, which affects the down-going trains more seri-
ously. The total delayed time is 1750 min and the objective value is 3250 min.
In the same manner, we did the data experiments with the typical fuzzy linear
programming model on Case 2. The re-scheduled timetables are shown in
Tables 7.11 and 7.12 and Fig. 7.7.

7.5.2 Analysis of the Computation Cases

According to the data in Tables 7.7 and 7.8, the summary delayed time of the
down-going trains is 590 min and the summary delayed time of up-going trains is
402.5 min. Compared to the results in Tables 7.5 and 7.6, the summary delayed
time of the down-going trains calculated out with the typical fuzzy linear pro-
gramming model is 105 min more than that with the improved fuzzy linear pro-
gramming model. Similarly, the summary delayed time of the up-going trains is
85 min more. Correspondingly, the optimal objective of the model is calculated to
Table 7.7 The re-scheduled timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the down-going direction with typical fuzzy linear
programming in computation Case 1
G507 G651 G501 G71 G653
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 8.0000 8.2100
Zhuozhou 8.2400 8.2400 8.4600 8.5400
Gaobeidian 8.2900 8.2900 8.5900 8.5900
Baoding 7.4600 7.4800 8.0900 8.0900 8.4100 8.4300 9.1100 9.1100
Dingzhou 7.5400 7.5600 8.0800 8.0800 8.2400 8.2600 8.5800 8.5800 9.2600 9.2600
Shijiazhuang 8.1900 8.2300 8.2400 8.2800 8.4900 8.5200 9.1900 9.2200 9.4600 9.4900
Gaoyi 8.3700 8.3700 8.4200 8.4200 9.0600 9.0600 9.3600 9.3600 10.0100 10.0100
7.5 Computation Cases and Analysis

Xingtai 8.5200 8.5200 8.5600 8.5800 9.2000 9.2200 9.5100 9.5100 10.1500 10.1500
Handan 9.0200 9.0400 9.1400 9.1600 9.3200 9.3200 10.0100 10.0300 10.2400 10.2400
Anyang 9.1800 9.1800 9.3000 9.3000 9.4700 9.4900 10.1700 10.1700 10.3700 10.3900
Hebi 9.3100 9.3300 9.4400 9.4600 10.0100 10.0100 10.3000 10.3000 10.5300 10.5300
Xinxiang 9.4900 9.5200 9.5700 9.5700 10.1100 10.1100 10.4100 10.4300 11.0500 11.1800
Zhengzhou 10.1300 10.1900 10.3100 11.0400 11.3900
G509 G83 G571 G511 G79
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 8.4300 9.0000 9.2700 9.3700 10.0000
Zhuozhou 9.0700 9.0700 9.3000 9.3000 10.0000 10.0000 10.1200 10.1400 10.3200 10.3200
Gaobeidian 10.1100 10.1100 9.3630 9.3630 10.0400 10.0400 10.1800 10.1800 10.3530 10.3530
Baoding 9.2400 9.2600 9.4600 9.4600 10.1600 10.1800 10.2900 10.2900 10.4400 10.4400
Dingzhou 9.4400 9.5100 9.5900 9.5900 10.3030 10.3030 10.4100 10.4100 10.5430 10.5430
Shijiazhuang 10.1400 10.2400 10.1900 10.2100 10.4800 10.5200 10.5800 11.0100 11.0900 11.1100
Gaoyi 10.3400 10.3400 10.3100 10.3100 11.0500 11.0500 11.1300 11.2500 11.2100 11.2100
(continued)
121
Table 7.7 (continued)
122

Xingtai 10.4700 10.4700 10.4130 10.4130 11.1800 11.2000 11.4230 11.4230 11.3200 11.3200
Handan 10.5600 10.5600 10.4900 10.4900 11.3600 11.4500 11.5500 11.5500 11.4030 11.4030
Anyang 11.0800 11.0800 10.5900 10.5900 11.5800 11.5800 11.5200 11.5200
Hebi 11.2000 11.2200 11.0900 11.0900
Xinxiang 11.3200 11.3200 11.1630 11.1630
Zhengzhou 11.5200 11.3200
G655 G513 G657 G515
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
7

Beijing 10.0500 10.4800 11.0600 11.5000

Zhuozhou 10.5000 10.5000 11.1300 11.1300 11.3000 11.3000
Gaobeidian 10.5500 10.5700 11.1800 11.2000 11.3500 11.3500
Baoding 11.0600 11.0800 11.3130 11.3130 11.4700 11.4900
Dingzhou 11.1830 11.1830 11.4500 11.4500
Shijiazhuang 11.3300 11.3600 12.0700 12.1100
Gaoyi 11.5100 11.5100
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
The time in italics is the rescheduled inbound and outbound time of the trains at stations
Train Re-scheduling Based on an Improved Fuzzy Linear …
Table 7.8 The re-scheduled timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the up-going direction with typical fuzzy linear
programming in computation Case 1
G560 G90 G508 G562 G652
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 12.0400 11.3300
Zhuozhou 11.3700 11.3900 11.1400 11.1400
Gaobeidian 11.2600 11.2800 11.1000 11.1000
Baoding 11.1000 11.1300 11.0100 11.0100 11.5230 11.5230 12.0600 12.0800
Dingzhou 10.5600 10.5600 10.4900 10.4900 11.3830 11.3830 11.5130 11.5130
Shijiazhuang 10.3700 10.3700 10.3100 10.3300 11.1500 11.1900 11.2800 11.3100 11.5700 12.0000
Gaoyi 10.2400 10.2400 10.1900 10.1900 11.0130 11.0130 11.1430 11.1430 11.4300 11.4500
7.5 Computation Cases and Analysis

Xingtai 10.1130 10.1130 10.0700 10.0700 10.4500 10.4700 11.0100 11.0100 11.3000 11.3000
Handan 9.5500 10.0100 9.5800 9.5800 10.3730 10.3730 10.4900 10.5100 11.2100 11.2100
Anyang 9.3000 9.3800 9.4600 9.4600 10.2500 10.2500 10.3300 10.3500 11.0800 11.0800
Hebi 9.1300 9.1500 9.3500 9.3500 10.1200 10.1400 10.2130 10.2130 10.5500 10.5700
Xinxiang 8.5500 8.5700 9.2600 9.2600 10.0330 10.0330 10.1000 10.1200 10.4400 10.4400
Zhengzhou 8.3500 9.0000 9.3200 9.4000 10.1400
G502 G654 G512 G672 G6732
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 10.2300
Zhuozhou 9.5900 9.5900
Gaobeidian 9.5430 9.5430
Baoding 9.3900 9.4200
Dingzhou 9.2400 9.2400
Shijiazhuang 9.0000 9.0300
Gaoyi 8.4300 8.4500
(continued)
123
Table 7.8 (continued)
124

Xingtai 8.1730 8.1730

Handan 12.0000 12.0000 8.0600
Anyang 11.5600 11.5600 11.4800 11.4800
Hebi 11.4400 11.4400 11.3830 11.3830
Xinxiang 11.2100 11.3400 11.2900 11.2900 11.5000 11.5200 12.0000 12.0200
Zhengzhou 11.0100 11.1200 11.3000 11.4000
G6734 G6704 G602 G92
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
7

Beijing 12.1900 8.5700 9.5300 11.0000

Zhuozhou 11.5200 11.5400 8.2900 8.3200 9.2900 9.2900 10.3800 10.3800
Gaobeidian 11.4100 11.4300 8.1800 8.2000 9.2500 9.2500 10.3400 10.3400
Baoding 11.2800 11.2800 8.0100 9.1000 9.1200 10.2100 10.2100
Dingzhou 11.1330 11.1330 8.5600 8.5600 10.0900 10.0900
Shijiazhuang 10.5100 10.5300 8.3400 9.5000
Gaoyi 10.3730 10.3730
Xingtai 10.2300 10.2300
Handan 10.1300
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
The time in italics is the rescheduled inbound and outbound time of the trains at stations
Train Re-scheduling Based on an Improved Fuzzy Linear …
7.5 Computation Cases and Analysis 125

Stations G655
G71 G653 G509 G83 G571 G511 G79 G513 G657 G515
Beijing

G501

Zhuozhou
Gaobeidian
G651
Baoding
G507

Dingzhou
G6704
Shijiazhuang
G92
G602
Gaoyi

Xingtai
Handan
G6732 G6734
Anyang

Hebi

Xinxiang

Zhengzhou
8:00 9:00 10:00 11:00 G654 G512 12:00
G560 G90 G508 G672 Time
G562 G652 G502

Fig. 7.5 The replanned train working diagram generated with the typical fuzzy linear
programming model from 8 to 12 a.m. in section between Beijing and Zhengzhou in computation
Case 1. The dotted lines stand for the original planned moving trajectories of the disturbed trains.
The red lines are the re-scheduled moving trajectories of the disturbed trains. The wavy lines imply
that the trains are disturbed when running in section between Beijing and Zhuozhou

be 4007.5, which is much smaller than 4197.5. We can conclude that the model
proposed in this chapter has more preeminent optimizing ability.
In Case 2, the optimal objective of the model is calculated to be 3183. The
summary delayed time of the down-going trains is 1684 min and the summary
delayed time of up-going trains is 133 min. Compared to the results in Tables 7.9
and 7.10, the summary delayed time of the down-going trains and the up-going
trains calculated out with the typical fuzzy linear programming model is both more
than that with the improved fuzzy linear programming model presented in this
chapter. It proves again that the model proposed in this chapter has more preemi-
nent optimizing ability.
To compare the computation efficiency of the improved fuzzy linear program-
ming and typical fuzzy linear programming, we recorded the computation time of
Table 7.9 The re-scheduled timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the down-going direction in computation Case 2
126

G507 G651 G501 G71 G653

Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 8.0000 8.2100
Zhuozhou 8.2400 8.2400 8.4600 8.5400
Gaobeidian 8.2900 8.2900 8.5900 8.5900
Baoding 7.4600 7.4800 8.0900 8.0900 8.4100 8.4300 9.1100 9.1100
Dingzhou 7.5400 7.5600 8.0800 8.0800 8.2400 8.2600 8.5800 8.5800 9.2600 9.2600
Shijiazhuang 8.1900 8.2300 8.2400 8.2800 8.4900 8.5200 9.1900 9.2200 9.4600 9.4900
7

Gaoyi 8.3700 8.3700 8.4200 8.4200 9.0600 9.0600 9.3600 9.3600 10.0100 10.0100
Xingtai 8.5200 8.5200 8.5600 8.5800 9.2000 9.2200 9.5100 9.5100 10.1500 10.1500
Handan 9.0200 9.0400 9.1400 9.1600 9.3200 9.3200 10.0100 10.0300 10.2400 10.2400
Anyang 9.1800 9.1800 9.3000 9.3000 9.4700 9.4900 10.1700 10.1700 10.3700 10.3900
Hebi 9.3100 9.3300 9.4400 9.4600 10.0100 10.0100 10.3000 10.3000 10.5300 10.5300
Xinxiang 9.4900 9.5200 9.5700 9.5700 10.1100 10.1100 10.4100 10.4300 11.0500 11.1800
Zhengzhou 10.1300 10.1900 10.3100 11.0400 11.3900
G509 G83 G571 G511 G79
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 9.0900 9.1200 9.2700 9.3700 10.2700
Zhuozhou 9.3030 9.3030 9.3530 9.3530 9.4630 9.4630 9.5800 10.0400 10.4830 10.4830
Gaobeidian 9.3500 9.3500 9.4000 9.4000 9.5200 9.5200 10.0830 10.0830 10.5300 10.5300
Baoding 9.4600 9.4800 9.5230 9.5230 10.0800 10.1000 10.2100 10.2100 11.0400 11.0400
Dingzhou 10.0200 10.0400 10.0630 10.0630 10.2500 10.2500 10.3600 10.3600 11.1600 11.1600
Shijiazhuang 10.2300 10.3100 10.2600 10.2800 10.4600 10.5000 10.5600 10.5900 11.3400 11.3600
Gaoyi 10.4200 10.4200 10.3900 10.3900 11.0330 11.0330 11.1300 11.2500 11.4600 11.4600
Xingtai 10.5330 10.5330 10.5030 10.5030 11.1800 11.2000 11.4230 11.4230 11.5630 11.5630
(continued)
Train Re-scheduling Based on an Improved Fuzzy Linear …
Table 7.9 (continued)
Handan 11.0130 11.0130 10.5830 10.5830 11.3600 11.4500 11.5500 11.5500 – –
Anyang 11.1230 11.1230 11.0930 11.0930 11.5800 11.5800 – –
Hebi 11.2400 11.2600 11.2100 11.2100
Xinxiang 11.3300 11.3300 11.2800 11.2800
Zhengzhou 11.5000 11.3500
G655 G513 G657 G515
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
Beijing 10.3000 11.3100 11.3400 11.5000
Zhuozhou 10.5200 10.5200 11.5200 11.5200 11.5500 11.5500
7.5 Computation Cases and Analysis

Gaobeidian 10.5600 10.5800 11.5600 11.5800 11.5900 11.5900

Baoding 11.0900 11.1100 – – – –
Dingzhou 11.2300 11.2300 – –
Shijiazhuang 11.3700 11.4000 – –
Gaoyi 11.5030 11.5030
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
The time in italics is the rescheduled inbound and outbound time of the trains at stations
127
Table 7.10 The re-scheduled timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the up-going direction in computation Case 2
128

G560 G90 G508 G562 G652

Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 11.3000
Zhuozhou 11.3700 – 11.0900 11.0900
Gaobeidian 11.2600 11.2800 11.0430 11.0430
Baoding 11.0700 11.1000 10.5300 10.5300 11.5100 11.5100
Dingzhou 10.5300 10.5300 10.4100 10.4100 11.3700 11.3700 11.5130 11.5130
Shijiazhuang 10.3400 10.3400 10.2100 10.2300 11.1300 11.1700 11.2800 11.3100 11.5500 11.5800
7

Gaoyi 10.2030 10.2030 10.0830 10.0830 11.0000 11.0000 11.1400 11.1400 11.3800 11.4000
Xingtai 10.0700 10.0700 9.5600 9.5600 10.4300 10.4500 10.5930 10.5930 11.2400 11.2400
Handan 9.5500 9.5700 9.4700 9.4700 10.3300 10.3300 10.4700 10.4900 11.1400 11.1400
Anyang 9.3000 9.3800 9.3400 9.3400 10.1830 10.1830 10.2800 10.3000 11.0000 11.0000
Hebi 9.1300 9.1500 9.2400 9.2400 10.0300 10.0500 10.1300 10.1300 10.4500 10.4700
Xinxiang 8.5500 8.5700 9.1600 9.1600 9.5330 9.5330 10.0000 10.0200 10.3400 10.3400
Zhengzhou 8.3500 9.0000 9.3200 9.4000 10.1400
G502 G654 G512 G672 G6732
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 10.2300
Zhuozhou 9.5900 9.5900
Gaobeidian 9.5430 9.5430
Baoding 9.3900 9.4200
Dingzhou 9.2400 9.2400
Shijiazhuang 9.0000 9.0300
Gaoyi 8.4300 8.4500
Xingtai 8.1730 8.1730
(continued)
Train Re-scheduling Based on an Improved Fuzzy Linear …
Table 7.10 (continued)
Handan 12.0000 12.0000 8.0600
Anyang 11.5600 11.5600 11.4800 11.4800
Hebi 11.4400 11.4400 11.3830 11.3830
Xinxiang 11.2100 11.3400 11.2900 11.2900 11.5000 11.5200 12.0000 12.0200
Zhengzhou 11.0100 11.1200 11.3000 11.4000
G6734 G6704 G602 G92
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
Beijing 12.1900 9.0800 10.2600 11.1500
Zhuozhou 11.5200 11.5400 8.2900 8.4700 9.2900 10.0500 10.3800 10.5300
7.5 Computation Cases and Analysis

Gaobeidian 11.4100 11.4300 8.1800 8.2000 9.2500 9.2500 10.3400 10.3400

Baoding 11.2800 11.2800 8.0100 9.1000 9.1200 10.2100 10.2100
Dingzhou 11.1330 11.1330 8.5600 8.5600 10.0900 10.0900
Shijiazhuang 10.5100 10.5300 8.3400 9.5000
Gaoyi 10.3730 10.3730
Xingtai 10.2300 10.2300
Handan 10.1300
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
The time in italics is the rescheduled inbound and outbound time of the trains at stations
129
130 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

G83
G509 G655 G655 G657
G71 G653 G509 G83 G571 G511 G79 G79 G513 G657 G513 G515
Beijing

G501

Zhuozhou
Gaobeidian
G651
Baoding
G507

Dingzhou
G6704
Shijiazhuang G92
G602
Gaoyi

Xingtai
Handan
G6732 G6734
Anyang

Hebi

Xinxiang

Zhengzhou
8:00 10:00 11:00 G654 G512 12:00
G560 9:00 G508
G90 G562 G652 G502 G672 Time

Fig. 7.6 The replanned train working diagram generated with the improved fuzzy linear
programming model from 8 to 12 a.m. in section between Beijing and Zhengzhou in computation
Case 2. The dotted lines stand for the original planned moving trajectories of the disturbed trains.
The red lines are the re-scheduled moving trajectories of the disturbed trains

the two algorithms when solving the train re-scheduling model in Case 1. We did
the data experiments 10 times with the two programming models respectively. The
time computation cost with the improved fuzzy linear programming varies from
1828 to 1837 ms, see Table 7.13. The average value is 1832.9 ms. The time cost
with typical fuzzy linear programming varies from 1650 to 1660 ms. The average
value is 1654.0 ms. The computation time cost with the typical linear programming
is 178.9 ms shorter that cost by the improved linear programming. It stems from the
fact that the improved fuzzy linear programming dealt with the boundaries of the
fuzzy coefficients, which cost the computation time. Even so, the improved fuzzy
programming is acceptable because of the computational performance.
We also recorded the computation time of the two algorithms when solving the
train re-scheduling model in Case 2. The average time cost with typical fuzzy linear
programming is 1332.6 ms, while it cost 1523.2 ms with the improved fuzzy linear
programming averagely. Case 2 also proved the improved linear programming is
considered acceptable.
From the computing results, we also conclude that the performance of the
proposed model on the two numerical examples is steady and robust because that
the cost time in the computations varies slightly in the two cases.
Table 7.11 The re-scheduled timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the down-going direction with typical fuzzy linear
programming in computation Case 2
G507 G651 G501 G71 G653
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 8.0000 8.2100
Zhuozhou 8.2400 8.2400 8.4600 8.5400
Gaobeidian 8.2900 8.2900 8.5900 8.5900
Baoding 7.4600 7.4800 8.0900 8.0900 8.4100 8.4300 9.1100 9.1100
Dingzhou 7.5400 7.5600 8.0800 8.0800 8.2400 8.2600 8.5800 8.5800 9.2600 9.2600
Shijiazhuang 8.1900 8.2300 8.2400 8.2800 8.4900 8.5200 9.1900 9.2200 9.4600 9.4900
Gaoyi 8.3700 8.3700 8.4200 8.4200 9.0600 9.0600 9.3600 9.3600 10.0100 10.0100
7.5 Computation Cases and Analysis

Xingtai 8.5200 8.5200 8.5600 8.5800 9.2000 9.2200 9.5100 9.5100 10.1500 10.1500
Handan 9.0200 9.0400 9.1400 9.1600 9.3200 9.3200 10.0100 10.0300 10.2400 10.2400
Anyang 9.1800 9.1800 9.3000 9.3000 9.4700 9.4900 10.1700 10.1700 10.3700 10.3900
Hebi 9.3100 9.3300 9.4400 9.4600 10.0100 10.0100 10.3000 10.3000 10.5300 10.5300
Xinxiang 9.4900 9.5200 9.5700 9.5700 10.1100 10.1100 10.4100 10.4300 11.0500 11.1800
Zhengzhou 10.1300 10.1900 10.3100 11.0400 11.3900
G509 G83 G571 G511 G79
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 9.0900 9.1200 9.2700 9.3700 10.2700
Zhuozhou 9.3030 9.3030 9.3530 9.3530 9.4630 9.4630 9.5900 10.0400 10.5000 10.5000
Gaobeidian 9.3500 9.3500 9.4000 9.4000 9.5200 9.5200 10.0830 10.0830 10.5430 10.5430
Baoding 9.4700 9.4900 9.5230 9.5230 10.0800 10.1000 10.2100 10.2100 11.0600 11.0600
Dingzhou 10.0300 10.0500 10.0800 10.0800 10.2500 10.2500 10.3600 10.3600 11.2000 11.2000
Shijiazhuang 10.2400 10.3200 10.2700 10.2900 10.4600 10.5000 10.5600 10.5900 11.3600 11.3800
Gaoyi 10.4200 10.4200 10.3900 10.3900 11.0330 11.0330 11.1300 11.2500 11.4800 11.4800
(continued)
131
Table 7.11 (continued)
132

Xingtai 10.5330 10.5330 10.5030 10.5030 11.1800 11.2000 11.4230 11.4230 11.5800 11.5800
Handan 11.0130 11.0130 10.5830 10.5830 11.3600 11.4500 11.5500 11.5500 – –
Anyang 11.1230 11.1230 11.0930 11.0930 11.5800 11.5800 – –
Hebi 11.2400 11.2600 11.2100 11.2100
Xinxiang 11.3300 11.3300 11.2800 11.2800
Zhengzhou 11.5000 11.3500
G655 G513 G657 G515
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
7

Beijing 10.3000 11.3100 11.3400 11.5000

Zhuozhou 10.5300 10.5300 11.5230 11.5230 11.5530 11.5530
Gaobeidian 10.5800 11.0000 11.5900 11.5900 12.0000 12.0000
Baoding 11.1100 11.1300 – – – –
Dingzhou 11.2400 11.2400 – –
Shijiazhuang 11.3700 11.4000 – –
Gaoyi 11.5030 11.5030
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
The time in italics is the rescheduled inbound and outbound time of the trains at stations
Train Re-scheduling Based on an Improved Fuzzy Linear …
Table 7.12 The re-scheduled timetable from 8 to 12 a.m. in section between Beijing and Zhengzhou in the up-going direction with typical fuzzy linear
programming in computation Case 2
G560 G90 G508 G562 G652
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 11.3000
Zhuozhou 11.3700 – 11.0900 11.0900
Gaobeidian 11.2600 11.2800 11.0430 11.0430
Baoding 11.0700 11.1000 10.5300 10.5300 11.5100 11.5100
Dingzhou 10.5300 10.5300 10.4100 10.4100 11.3700 11.3700 11.5130 11.5130
Shijiazhuang 10.3400 10.3400 10.2100 10.2300 11.1300 11.1700 11.2800 11.3100 11.5500 11.5800
Gaoyi 10.2030 10.2030 10.0830 10.0830 11.0000 11.0000 11.1400 11.1400 11.3800 11.4000
7.5 Computation Cases and Analysis

Xingtai 10.0700 10.0700 9.5600 9.5600 10.4300 10.4500 10.5930 10.5930 11.2400 11.2400
Handan 9.5500 9.5700 9.4700 9.4700 10.3300 10.3300 10.4700 10.4900 11.1400 11.1400
Anyang 9.3000 9.3800 9.3400 9.3400 10.1830 10.1830 10.2800 10.3000 11.0000 11.0000
Hebi 9.1300 9.1500 9.2400 9.2400 10.0300 10.0500 10.1300 10.1300 10.4500 10.4700
Xinxiang 8.5500 8.5700 9.1600 9.1600 9.5330 9.5330 10.0000 10.0200 10.3400 10.3400
Zhengzhou 8.3500 9.0000 9.3200 9.4000 10.1400
G502 G654 G512 G672 G6732
Arrive Depart Arrive Depart Arrive Depart Arrive Depart Arrive Depart
Beijing 10.2300
Zhuozhou 9.5900 9.5900
Gaobeidian 9.5430 9.5430
Baoding 9.3900 9.4200
Dingzhou 9.2400 9.2400
Shijiazhuang 9.0000 9.0300
Gaoyi 8.4300 8.4500
(continued)
133
Table 7.12 (continued)
134

Xingtai 8.1730 8.1730

Handan 12.0000 12.0000 8.0600
Anyang 11.5600 11.5600 11.4800 11.4800
Hebi 11.4400 11.4400 11.3830 11.3830
Xinxiang 11.2100 11.3400 11.2900 11.2900 11.5000 11.5200 12.0000 12.0200
Zhengzhou 11.0100 11.1200 11.3000 11.4000
G6734 G6704 G602 G92
Arrive Depart Arrive Depart Arrive Depart Arrive Arrive
7

Beijing 12.1900 9.0900 10.2800 11.1700

Zhuozhou 11.5200 11.5400 8.2900 8.4800 9.2900 10.0500 10.3800 10.5500
Gaobeidian 11.4100 11.4300 8.1800 8.2000 9.2500 9.2500 10.3400 10.3400
Baoding 11.2800 11.2800 8.0100 9.1000 9.1200 10.2100 10.2100
Dingzhou 11.1330 11.1330 8.5600 8.5600 10.0900 10.0900
Shijiazhuang 10.5100 10.5300 8.3400 9.5000
Gaoyi 10.3730 10.3730
Xingtai 10.2300 10.2300
Handan 10.1300
note aa.bbcc stands for bb minutes cc seconds at aa o’clock
The time in italics is the rescheduled inbound and outbound time of the trains at stations
Train Re-scheduling Based on an Improved Fuzzy Linear …
7.5 Computation Cases and Analysis 135

G83
G509 G655 G655 G657
G71 G653 G509 G83 G571 G511 G79 G79 G513 G657 G513 G515
Beijing

G501

Zhuozhou
Gaobeidian
G651
Baoding
G507

Dingzhou
G6704
Shijiazhuang G92
G602
Gaoyi

Xingtai
Handan
G6732 G6734
Anyang

Hebi

Xinxiang

Zhengzhou
8:00 10:00 11:00 G654 G512 12:00
G560 9:00 G508
G90 G562 G652 G502 G672 Time

Fig. 7.7 The replanned train working diagram generated with the typical fuzzy linear
programming model from 8 to 12 a.m. in section between Beijing and Zhengzhou in computation
Case 2. The dotted lines stand for the original planned moving trajectories of the disturbed trains.
The red lines are the re-scheduled moving trajectories of the disturbed trains

Table 7.13 The computation time in the two cases

Case 1 Case 2
With improved With typical fuzzy With improved With typical fuzzy
fuzzy linear linear fuzzy linear linear
programming programming programming programming
1 1832 1654 1524 1331
2 1836 1653 1526 1332
3 1834 1650 1530 1333
4 1830 1654 1520 1332
5 1831 1658 1524 1334
6 1828 1660 1521 1332
7 1837 1650 1524 1331
8 1832 1654 1521 1335
9 1834 1652 1522 1332
10 1835 1654 1520 1334
Average 1832.9 1654.0 1523.2 1332.6
136 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

7.6 Brief Summary

On the operational planning level, a railway train re-scheduling problem is inves-

tigated under the uncertain environment of fuzziness. In the problem, the coeffi-
cients of the resources, which are on the right side of the constraints, are supposed
to have the fuzzy boundary value ranges. For this case, the traditional linear fuzzy
programming model will turn meaningless, and we improve the model, describing
the boundaries of the coefficients as fuzzy numbers.
On the basis of the improved fuzzy linear programming, the train re-scheduling
problem with fuzzy constraints is studied, which belonged to the operational level
of railway operation. For the convenience of solving models, some coefficients on
the right side of the constraints equation were simplified. The train re-scheduling
model was turned to a parameter linear programming model with the triangle
membership function. Two computation cases in different scenarios are listed and
used to verify the model. The numerical examples show that the designed algorithm
is steady and robust for not very large-scale problems.
Additionally, it is worth pointing out that the main focus of this chapter is to
provide the different decision-making methods for train re-scheduling problem
under the fuzzy environment. Generally, it is not easy to determine which model is
the best, and the applications of models are dependent on decision-makers’ pref-
erences. The approach to re-scheduling trains can help the dispatchers to redesign
the high-quality timetable.

References

Acuna-Agost, R., Michelon, P., Feillet, D., & Gueye, S. (2011a). A MIP-based local search
method for the railway re-scheduling problem. Networks, 57(1), 69–86.
Acuna-Agost, R., Michelon, P., Feillet, D., & Gueye, S. (2011b). SAPI: statical analysis of
propagation of incidents. A new approach for re-scheduling trains after disruptions. European
Journal of Operational Research, 215(1), 227–243.
Adenso-Díaz, B., González, M. O., & González-Torre, P. (1999). On-line timetable re-scheduling
in regional train services. Transportation Research Part B, 33(6), 387–398.
Cacchiani, V., Caprara, A., & Toth, P. (2010). Scheduling extra freight trains on railway networks.
Transportation Research Part B, 44(2), 215–231.
Castillo, E., Gallego, I., Ureña, J. M., & Coronado, J. M. (2011). Timetabling optimization of a
mixed double-and single-tracked railway network. Applied Mathematical Modelling, 35(2),
859–878.
Chang, C. S., & Kwan, C. M. (2005). Evaluation of evolutionary algorithms for multi-objective
train schedule optimization. Lecture Notes in Computer Science, 3339(1), 803–815.
Cordeau, J., Toth, P., & Vigo, D. (1998). A survey of optimization models for train routing and
scheduling. Transportation Science, 32(4), 380–404.
D’Ariano, A., Pacciarelli, D., & Pranzo, Macro. A. (2007). branch and bound algorithm for
scheduling trains in a railway network. European Journal of Operational Research, 183(2),
643–657.
References 137

Delgado, M., Herrera, F., Verdegay, J. L., & Vila, M. A. (1993). Post-optimality analysis on the
membership functions of a fuzzy linear programming problem. Fuzzy Sets and Systems, 53(3),
289–297.
Dorfman, M. J., & Medanic, J. (2004). Scheduling trains on a railway network using a discrete
event model of railway traffic. Transportation Research Part B, 38(1), 81–98.
Dubey, D., Chandra, S., & Mehra, A. (2012). Fuzzy linear programming under interval uncertainty
based on IFS representation. Fuzzy Sets and Systems, 188(1), 68–87.
Dubey, D., & Mehra, A. (2014). A bipolar approach in fuzzy multi-objective linear programming.
Fuzzy Sets and Systems, 246, 127–141.
Dündar, S., & Şahin, İ. (2012). Train re-scheduling with genetic algorithms and artificial neural
networks for single-track railways. Transportation Research Part C, 27, 1–15.
Ebrahimnejad, A. (2011). Sensitivity analysis in fuzzy number linear programming problems.
Mathematical and computer modeling, 53(9–10), 1878–1888.
Ebrahimnejad, A., & Tavana, M. (2014). A novel method for solving linear programming
problems with symmetric trapezoidal fuzzy numbers. Applied Mathematical Modelling, 38(17–
18), 4388–4395.
Fan, Y. R., Huang, G. H., & Yang, A. L. (2013). Generalized fuzzy linear programming for
decision making under uncertainty: Feasibility of fuzzy solutions and solving approach.
Information Sciences, 241, 12–27.
Frank, R., Neggers, J., & Jun, Y. (2008). Method for optimizing linear problems with fuzzy
constraints. International Mathematical Forum, 3(21–24), 1141–1155.
Gasimov, R. N., & Yenilmez, K. (2002). Solving fuzzy linear programming problems with linear
membership functions. Turkish Journal of Mathematics, 26(4), 375–396.
Hajiagha, S. H. R., Mahdiraji, H. A., & Hashemi, S. S. (2013). Multi-objective linear
programming with interval coefficients: A fuzzy set based approach. Kybernetes, 42(3), 482–
496.
Jin, L., Huang, G. H., Cong, D., & Fan, Y. R. (2014). A Robust Inexact Joint-optimal a cut
Interval Type-2 Fuzzy Boundary Linear Programming (RIJ-IT2FBLP) for energy systems
planning under uncertainty. International Journal of Electrical Power & Energy Systems, 56,
19–32.
Kaur, J., & Kumar, A. (2013). Mehar’s method for solving fully fuzzy linear programming
problems with L-R fuzzy parameters. Applied Mathematical Modelling, 37(12–13), 7142–
7153.
Kazuo, N. (1984). Some extensions of fuzzy linear programming. Fuzzy Sets and Systems, 14(3),
211–229.
Krasemann, J. T. (2012). Design of an effective algorithm for fast response to the re-scheduling of
railway traffic during disturbances. Transportation Research Part C, 20(1), 62–78.
Kroon, L., Huisman, D., Abbink, E., Fioole, P. J., & Fischetti, M. (2009). The new Dutch
timetable: The OR revolution. Interfaces, 39(1), 6–17.
Kroon, L., Romeijn, H., & Zwaneweld, P. J. (1997). Routing trains through railway networks:
Complexity issues. European Journal of Operational Research, 98, 485–498.
Kumar, A., & Kaur, J. (2013). General form of linear programming problems with fuzzy
parameters. Journal of Applied Research and Technology, 5, 629–635.
Mazzarello, M., & Ottaviani, E. (2007). A traffic management system for real-time traffic
optimization in railways. Transportation Research Part B, 41(2), 246–274.
Meng, X., Jia, L., Qin Y., & Xu, J. (2013). Hybrid timed event graph model for networked train
operation simulation and timetable stability optimization. In Proceedings of the 2013
International Conference on Electrical and Information Technologies for Rail Transportation
(EITRT2013) (Vol. I, Changchun, China, pp. 575–582). 2014. Berlin Heidelberg: Springer.
Min, Y. H., Park, M. J., Hong, S. P., & Hong, S. H. (2011). An appraisal of a
column-generation-based algorithm for centralized train-conflict resolution on a metropolitan
railway network. Transportation Research Part B, 45(2), 409–429.
Narayanaswami, S., & Rangaraj, N. (2011). Scheduling and re-scheduling of railway operations: A
review and expository analysis. Technology Operation Management, 2(2), 102–122.
138 7 Train Re-scheduling Based on an Improved Fuzzy Linear …

Rena, A., & Wanga, Y. (2014). Optimistic Stackelberg solutions to bi-level linear programming
with fuzzy random variable coefficients. Knowledge-Based Systems, 67, 206–217.
Rodriguez, J. (2007). Constraint programming model for real-time trains scheduling at junctions.
Transportation Research Part B, 41(2), 231–245.
Şahin, İ. (1999). Railway traffic control and train scheduling based on inter-train conflict
management. Transportation Research Part B, 33(7), 511–534.
Sakawa, M., & Matsui, T. (2013). Interactive fuzzy random cooperative two-level linear
programming through level sets based probability maximization.Expert Systems with
Application, 40(4), 1400–1406.
Schobel, A. (2001). A model for the delay management problem based on
mixed-integer-programming. Electronic Notes in Theoretical Computer Science, 50(1), 1–10.
Tanaka, H. (1984). Fuzzy linear programming problems with fuzzy numbers. Fuzzy Sets and
Systems, 13, 1–10.
Tőrnquist, J., & Persson, J. A. (2007). N-tracked railway traffic re-scheduling during disturbances.
Transportation Research Part B, 41(3), 342–362.
Wan, S., & Dong, J. (2014). Possibility linear programming with trapezoidal fuzzy numbers.
Applied Mathematical Modelling, 38(5–6), 1660–1672.
Yang, L., Gao, Z., & Li, K. (2011). Railway freight transportation planning with mixed uncertainty
of randomness and fuzziness. Applied Soft Computing, 11(1), 778–792.
Yano, H., & Matsui, K. (2013). Hierarchical multiobjective fuzzy random linear programming
problems. Procedia Computer Science, 22, 162–171.