Intelligent Control (0640734)

Lecture (1)
Intelligent Control Systems: An Introduction

Prof. Kasim M. Al-Aubidy

Philadelphia University-Jordan

IntConSys-MSc Prof. Kasim M. Al-Aubidy 1

Course Title: Intelligent Control Systems
Instructor: Prof. Kasim M. Al-Aubidy.
Email: [email protected]
Semester: First, 2020-2021
Time: (12:00-15:00) Saturday.
Office Hours: (13:00-15:00) Wednesday.
Appointments to discuss the course should be made by email.

Course Material:
http://www.philadelphia.edu.jo/academics/kaubaidy/page.php?id=7

Course Description:
Industrial systems are complex and have nonlinear behaviors, therefore, new methodologies are required to
design and develop intelligent controllers. Intelligent control systems are becoming very important for both
academia and industry. Control methodologies are required to improve the performance of control complex
and nonlinear systems. These controller are based on soft computing tools such are fuzzy logic, neural network
and evolutionary computation.
IntConSys-MSc Prof. Kasim M. Al-Aubidy 2
Lectures: 45 hours/semester, 3 hours/week.
Homework and assignments: 4
Semester project: 2 projects for each student .

Course Objectives:
The main objective of this course is to :
 Provide a general introduction to intelligent systems .
 Provide examples of rule-based control systems.
 Describe design requirements of intelligent controllers.
 Study a range of methodologies for specifying and designing intelligent systems.
 Understand control methodologies developed using soft computing tools such as fuzzy logic, neural nets and
GAs.
 Describe and apply systems engineering methods and techniques in the design and analysis of intelligent
control systems for mechatronics applications.
 The course will involve several design projects.
Prerequisites:
Students are expected to be familiar with control systems, artificial intelligence, systems design and
implementation, programming with machine language and C++, systems modeling and simulation techniques.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 3

Intended Learning Outcomes:
On completing the course, students will enable to have to following skills:
Knowledge and understanding:
A1. Know the advantages and drawbacks of intelligent controllers and therefore when to apply them
A2. Understand how to derive, develop, and apply intelligent controllers
Intellectual skills:
B1 Comprehend advanced mathematical models and intelligent systems
B2. Design intelligent systems for various applications.
Professional and practical skills
C1. Simulate and analyze responses to advanced controller concepts
C2. Apply intelligent controllers to physical systems
General and transferrable skills
D1. Apply intelligent decision making techniques to engineering systems
D2. Optimize system performance.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 4

Grades:
Assignments, Project & Presentation: 20%,
Mid Exam: 30%
Final Exam: 50%

Projects:
 Define your own project and write a proposal.
 Experimental investigation requires a programming project and a final report.
 Final report contents: Project title, Objective, Introduction, Hardware design, Software design, Conclusion,
References.
 Team projects are allowed, but they must be significant!

IntConSys-MSc Prof. Kasim M. Al-Aubidy 5

Timetable:
Week Material to be covered Task
1 An introduction to classical and intelligent control systems.
2 Intelligent systems and applied artificial intelligence.
3 Artificial neural networks: fundamentals. HW1
4 Intelligent control concepts. Project Selection
5 Artificial neural networks: architectures Assignment1
6 Artificial neural networks: applications. Project (Phase1)
7 Introduction to fuzzy logic. Mid Exam
8 Fuzzy control and stability. HW2
9 Control applications of fuzzy logic. Assignment2

IntConSys-MSc Prof. Kasim M. Al-Aubidy 6

Timetable:
Week Material to be covered Task
10 Neuro-fuzzy controller: theory and design.
11 Neuro-fuzzy controller: applications. Assignment3
12 Probabilistic and evolutionary algorithms.
13 Optimization of intelligent systems using GA. Project (Phase2)
14 Intelligent control systems: research paper analysis
15 Intelligent control systems: design methods. Project (Phase3)
16 Mini Projects Final Exam

IntConSys-MSc Prof. Kasim M. Al-Aubidy 7

Text Books:
1. Intelligent Control Systems Using Soft Computing Methodologies,
By: Ali Zilouchian & Mo Jamshidi, CRC Press, 2001, ISBN:0-8493-
1875-0. Available online:
https://b-ok.asia/book/593496/3c6d6e?regionChanged=&redirect=7850341
 This book comprises most of the lecture notes for the course and is required
reading for all students.
 All selected material in this book is examinable.

2. Intelligent Control Systems with LabVIEW, By: Pedro Ponce-Cruz,

Fernando D. Ramírez-Figueroa, Springer, 2010, ISBN 978-1-84882-
683-0. Available online:
https://b-ok.asia/book/812190/493ebf?regionChanged=&redirect=7855146
 This book is optional, but provides further detail on the practical aspects of the
course.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 8

Why Intelligent Control Systems?
 Control systems theory has always been at the heart of Mechatronics.
 Many control methods have been developed, each has its own advantages and
disadvantages.
 Control methods can be divided into two categories:
 Classical Control; obtain a model of the plant to be controlled!
 Intelligent Control; does not need to know all about the plant to be controlled.
 Mechatronics are complex systems.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 9

Distinguishing between Classical and Intelligent Control:
 The mathematical approach is used in developing a classical control system, the designer
must mathematically model the plant.
 In developing an intelligent control, the designer doesn’t need to know the model of the
plant to be controlled, only the plant behavior is required. In many cases, the plant may be
too complex to be modeled.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 10

Classical and Intelligent Control:
 For classical control, the designer must model the plant and therefore the intelligence (or
knowledge) lies with him. The intelligence is shifted towards the designer.
 For intelligent control, the software abstractly models the plant and therefore the
intelligence (or knowledge) lies with it. The intelligence is shifted towards the software.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 11

Classical and Intelligent Control:
Classical Control Intelligent Control
Mathematical Modeling - Designer Abstract Modeling - Designer input the
Basic Concept designed the system which includes behavior to the system and then system
system dynamics attempt to abstractly define the system
Need to know prior information about Does not need to know all about the
the system dynamics system dynamics and conditions
Characteristics Suitable for system that can be easily
Appropriate for complex system
model
Open loop system Fuzzy logic
Examples of Closed loop system Artificial Neural Network
System Modeling Genetic Algorithm
Methods
Support vector machine
Swarm Intelligence
Particle Intelligence

IntConSys-MSc Prof. Kasim M. Al-Aubidy 12

Classical Control Methodology:
Solve the control problem by doing the following steps:
 Develop a model of the plant to be controlled.
 Design a controller using the mathematical model.
 Study plant performance using the mathematical model of the closed loop system.
 Implement the controller and evaluate the performance of the closed loop system (again,
possibly leading to redesign).

IntConSys-MSc Prof. Kasim M. Al-Aubidy 13

Methods for Classical Control:
Open-Loop Control System:
A signal is sent to a plant in order to make it move to a certain position. There is no relationship
or feedback from the plant to ensure that it went to the desired position.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 14

Methods for Classical Control:
Closed-Loop Control System:
The difference between the desired position and actual position is used by the controller to
generate a control signal to the plant. There is feedback from the plant to ensure that it went
to the desired position.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 15

Methods for Intelligent Control:
No System Modeling:
The designer only has to input the appropriate stimuli to the intelligent control and evaluate
it on its output.
The intelligent control itself develops a model of the system to be controlled.
Humans can perform complex tasks without knowing exactly how they do them.
The types of intelligent control includes: fuzzy logic, artificial neural networks, genetic
programming, and others.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 16

Definition of intelligent systems:
Intelligence is a mental quality that consists of the abilities to;
 learn from experience,
 adapt to new situations,
 understand and handle abstract concepts, and
 use knowledge to manipulate one’s environment.

Artificial Intelligence (AI):

AI is defined as “the ability of a digital computer or computer-controlled robot to perform
tasks commonly associated with intelligent beings”.
 The definition of intelligent systems is a function of expectations and the status of the
present knowledge: perhaps the “intelligent systems” of today are the “classical
systems” of tomorrow.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 17

Definition of intelligent systems:
 During the last four decades, researchers have proposed many model-based control
strategies that involve various phases such as modeling, analysis, simulation,
implementation and verification.
 Zadeh has considered this decade as the era of intelligent systems
“I believe the system analysis and controls should embrace soft computing and assign a
higher priority to the development of methods that can cope with imprecision,
uncertainties and partial truth.”
 Intelligent control is designed to seek control methods that provide a level of intelligence
and autonomy in the control decision that allows for improving the system performance.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 18

Soft computing & Hard computing:
 In hard computing, the prime objectives of the computations are precision and certainty.
 In soft computing, the precision and certainty carry a cost.
 Therefore, it is realistic to consider the integration of computation, reasoning, and
decision making in order to provide a framework for the trade off between precision and
uncertainty.
 The principal partners in such a consortium are fuzzy logic, neural network computing,
generic algorithms and probabilistic reasoning as well as their integration.
 Intelligent systems underlie what is called “soft computing.”

IntConSys-MSc Prof. Kasim M. Al-Aubidy 19

Soft computing Tools:
 Soft computing tools are commonly used to enhance AI and incorporate human expert
knowledge in computing processes.
 Their applications include the design of intelligent autonomous systems (controllers) and
dealing with unknown parameters.
 Soft computing can:
 learn from experience,
 universalize into domains where direct experience is absent,
 can perform mapping from inputs to the outputs faster than serial analytical
representations by using parallel computer architectures that simulate biological
processes.
 Intelligent control techniques that emulate characteristics of biological systems offer
opportunities for creating control products with new capabilities.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 20

Soft computing & Hard computing:
 Intelligent control has different tools for emulating the biological behavior that could solve
problems as human beings do. The main tools for IC are;
 Fuzzy logic systems are based on the experience of a human operator, expressed in a
linguistic form (IF–THEN rules).
 Artificial neural networks emulate the learning process of biologic neural networks, so
that the network can learn different patterns using a training method, supervised or
unsupervised.
 Evolutionary methods are based on evolutionary processes such as natural evolution.
These are essentially optimization procedures.
 Predictive methods are mathematical methods that provide information about the future
system behavior.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 21

Design of Intelligent Control Systems:
 The study of intelligent control systems requires;
 defining some important expressions that clarify these systems,
 understanding the desired application goals, and also
 understanding different tools of soft computing.

 Several software development platforms are used for developing intelligent control
systems. The LabVIEW is one of the most important software platforms used by
researchers for developing engineering applications and could be connected with different
hardware systems, as well as running standalone programs for simulating the controller’s
performance (validating the controller by simulation then implementing it). In addition,
LabVIEW is a graphical program that is very easy to learn.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 22

Soft computing Tools:
Neural Networks:
 For many decades, it has been a goal of engineers and scientists to develop a machine
with simple elements similar to one found in the human brain.
 Artificial neural networks (ANNs) are composed of many simple elements emulating
various brain activities and operating in parallel. The network function is determined
largely by the connections between these elements. Neural networks can be trained to
perform complex functions due to the nature of their nonlinear mappings of input to
output data set.
 Recently, the NN has been applied successfully to many fields of Engineering.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 23

Soft computing Tools:
Fuzzy Logic:
 The first paper in fuzzy set theory was written by Zadeh in 1965. In this paper he was
implicitly advancing the concept of human approximate reasoning to make effective
decisions on the basis of available imprecise linguistic information.
 The first implementation of Zadeh’s idea was accomplished in 1975 by Mamdani, and
demonstrated the viability of fuzzy logic control for a small model steam engine.
 After this pioneer work, many consumer products as well as other high tech applications
using fuzzy technology have been developed.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 24

Why Fuzzy Logic Control?
 FL provides a framework for both information and knowledge-based systems. So called
knowledge-based methodology is much closer to human thinking and natural language
than the traditionally classical logic.
 FL controller utilizes fuzzy logic to convert the linguistic control strategy based on expert
knowledge into an automatic control strategy.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 25

Soft computing Tools:
Evolutionary Computation:
There are various approaches to evolutionary optimization algorithms including evolution
concept, genetic programming and genetic algorithms. These algorithms are similar in their
basic concepts of evolution and differ in their approach to parameter representation.

Genetic Algorithm (GA) is an evolutionary algorithm which has performed well in noisy,
nonlinear and uncertain processes. GAs are desirable tools for optimizing problems with
exceeding complexity.

Genetic Programming (GP) is a symbolic-based nonlinear optimization. The GP paradigm

computationally simulates the Darwinian evolution process by applying fitness-based selection
and genetic operators to a population of parse trees of a given programming language. It
departs from the conventional GA primarily with regard to its representation scheme.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 26

Soft computing Tools: Hybrid Systems:
 Fuzzy logic, neural networks and evolutionary computations are complementary
methodologies in the design and implementation of IC systems.
 Each approach has its merits and drawbacks. To take advantage of the merits and eliminate
their drawbacks, several integration of these methodologies have been proposed;
1. Neuro-fuzzy systems: provide the fuzzy systems with automatic tuning systems using
NN as a tool.
2. Fuzzy neural network: retain the functions of NN with fuzzification of some of their
elements. For instance, fuzzy logic can be used to determine the learning steps of NN
structure.
3. Fuzzy-neural hybrid systems: utilize both fuzzy logic and neural networks in a system
to perform separate tasks for decouple subsystems.
4. GP-Fuzzy System, a population comprising fuzzy rule-bases (symbolic structures) that
are candidate solutions to the problem, evolves in response to selective pressure
induced by their relative success at implementing the desired behavior.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 27

Controller Design Constraints:
Cost: What is the total cost to implement the controller?
Time: How much time will it take to develop the controller?
Complexity: How much computation power and memory size will it take to implement the
controller?
Manufacturability: Does the designed controller need special requirements?
Reliability: What is its “mean time between failures?”
Adaptability: Can the proposed design be adapted to other similar applications.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 28

References:
1. Gupta, M. and Sinha, N. (eds.), Intelligent Control Systems: Theory and Applications, IEEE
Press, Piscataway, NJ, 1996.
2. Lime, C.M. and Hiyama, T., Application of Fuzzy Control to a Manipulator, IEEE Trans. on
Robotics and Automation, Vol. 7, 5,1991.
3. Li, W., Neuro-Fuzzy Systems for Intelligent Robot Navigation and Control Under
Uncertainty, Proc. of IEEE Robotics and Automation Conf., 1995.
4. Nedungadi, A., Application of Fuzzy Logic to Solve the Robot Inverse Kinematic Problem,
Proc. of Fourth World Conf. on Robotics Research, 1, 1991.
5. Jang, J., Sun, C., and Mizutani, E., Neuro Fuzzy and Soft Computing, Prentice Hall, Upper
Saddle River, NJ, 1997.

IntConSys-MSc Prof. Kasim M. Al-Aubidy 29