Operations Research PDF

CONTENTS
1. INTRODUCTION TO OR..................................................................................... 1
1.1 TERMINOLOGY ...................................................................................................... 1
1.2 THE METHODOLOGY OF OR ............................................................................. 1
1.3 HISTORY OF OR .................................................................................................... 2
2. BASIC OR CONCEPTS ...................................................................................... 6
3. LINEAR PROGRAMMING ................................................................................ 11
OPERATIONS RESEARCH 3.1 FORMULATING LP............................................................................................... 13
LECTURE NOTES 3.1.1 Giapetto Example .......................................................................................... 13
3.1.2 Advertisement Example................................................................................ 15
3.1.3 Diet Example .................................................................................................. 16
3.1.4 Post Office Example...................................................................................... 17
3.1.5 Sailco Example .............................................................................................. 18
3.1.6 Customer Service Level Example............................................................... 19
3.2 SOLVING LP .......................................................................................................... 20
Y. !lker Topcu, Ph.D. 3.2.1 LP Solutions: Four Cases............................................................................. 20
3.2.2 The Graphical Solution (Minimization) ....................................................... 21
3.2.3 The Graphical Solution (Maximization) ...................................................... 22
3.2.4 The Simplex Algorithm.................................................................................. 23
3.2.5 Example illustrating Simplex (Dakota Furniture) ...................................... 24
3.2.6 Other Solution Cases by Simplex ............................................................... 29
3.2.7 The Big M Method ......................................................................................... 31
3.2.8 Example illustrating Big M (Oranj Juice).................................................... 32
3.3 SENSITIVITY ANALYSIS..................................................................................... 35
3.3.1 Reduced Cost................................................................................................. 35
3.3.2 Shadow Price ................................................................................................. 35
3.3.3 Simple Example ............................................................................................. 36
3.3.4 Utilizing Lindo Output for Sensitivity ........................................................... 37
Acknowledgements:
We would like to acknowledge Prof. W.L. Winston's "Operations Research: Applications and 3.3.5 Additional Example (Utilizing Simplex for Sensitivity).............................. 39
Algorithms" and Prof. J.E. Beasley's lecture notes which greatly influence these notes...
We retain responsibility for all errors and would love to hear from visitors of this site!
3.4 DUALITY ................................................................................................................. 42
Istanbul Technical University OR/MS team 3.4.1 The Dual Theorem......................................................................................... 42
www.isl.itu.edu.tr/ya 3.4.2 Finding the Dual of a Normal Max Problem .............................................. 42
3.4.3 Finding the Dual of a Normal Min Problem ............................................... 43
Dr. Y. !lker Topcu (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

i
3.4.4 Finding the Dual of a Nonnormal Max Problem........................................ 43
3.4.5 Finding the Dual of a Nonnormal Min Problem......................................... 43
3.4.6 An Example for Economic Interpretation (Dakota Furniture).................. 44
4. TRANSPORTATION PROBLEMS.................................................................... 45 1. INTRODUCTION TO OR
4.1 FORMULATING TRANSPORTATION PROBLEMS ....................................... 46
4.1.1 LP Representation......................................................................................... 46 1.1 TERMINOLOGY
4.1.2 Powerco Example for Balanced Transportation Problem ....................... 46 The British/Europeans refer to "operational research", the Americans to "operations
4.1.3 Balancing an Unbalanced Transportation Problem ................................. 47 research" - but both are often shortened to just "OR" (which is the term we will use).
4.1.4 Modified Powerco Example for Excess Supply ........................................ 48 Another term which is used for this field is "management science" ("MS"). The
4.1.5 Modified Powerco Example for Unmet Demand....................................... 49 Americans sometimes combine the terms OR and MS together and say "OR/MS" or
4.2 FINDING BFS FOR TRANSPORT’N PROBLEMS .......................................... 50 "ORMS".
4.2.1 Northwest Corner Method ............................................................................ 51 Yet other terms sometimes used are "industrial engineering" ("IE"), "decision
4.2.2 Minimum Cost Method.................................................................................. 53 science" ("DS"), and “problem solving”.
4.2.3 Vogel’s Method .............................................................................................. 55 In recent years there has been a move towards a standardization upon a single term
4.3 THE TRANSPORTATION SIMPLEX METHOD ............................................... 57 for the field, namely the term "OR".
4.3.1 Steps of the Method ...................................................................................... 57

4.3.2 Solving Powerco Example by Transportation Simplex............................ 58 1.2 THE METHODOLOGY OF OR
4.4 TRANSSHIPMENT PROBLEMS ........................................................................ 61 When OR is used to solve a problem of an organization, the following seven step
4.4.1 Steps................................................................................................................ 61 procedure should be followed:
4.4.2 Kuruoglu Example ......................................................................................... 62 Step 1. Formulate the Problem
4.5 ASSIGNMENT PROBLEMS ................................................................................ 64 OR analyst first defines the organization's problem. Defining the problem includes
4.5.1 LP Representation......................................................................................... 64 specifying the organization's objectives and the parts of the organization (or system)
4.5.2 Hungarian Method ......................................................................................... 64 that must be studied before the problem can be solved.
4.5.3 Flight Crew Example ..................................................................................... 66 Step 2. Observe the System

Next, the analyst collects data to estimate the values of parameters that affect the
5. INTEGER PROGRAMMING ............................................................................. 68
organization's problem. These estimates are used to develop (in Step 3) and
5.1 FORMULATING IP................................................................................................ 69
evaluate (in Step 4) a mathematical model of the organization's problem.
5.1.1 Budgeting Problems...................................................................................... 69
Step 3. Formulate a Mathematical Model of the Problem
5.1.2 Knapsack Problems ...................................................................................... 72
The analyst, then, develops a mathematical model (in other words an idealized
5.1.3 Fixed Charge Problems................................................................................ 73
representation) of the problem. In this class, we describe many mathematical
5.1.4 Set Covering Problems................................................................................. 78
techniques that can be used to model systems.
5.1.5 Either - Or Constraints .................................................................................. 80
5.1.6 If - Then Constraints...................................................................................... 81
5.1.7 Traveling Salesperson Problems ................................................................ 82
Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

ii 1
Step 4. Verify the Model and Use the Model for Prediction experiments for both the Air Force and the Army would be carried out. Experimental
The analyst now tries to determine if the mathematical model developed in Step 3 is radar equipment was brought up to a high state of reliability and ranges of over 100
an accurate representation of reality. To determine how well the model fits reality, miles on aircraft were obtained.
one determines how valid the model is for the current situation. It was also in 1936 that Royal Air Force (RAF) Fighter Command, charged
Step 5. Select a Suitable Alternative specifically with the air defense of Britain, was first created. It lacked however any
Given a model and a set of alternatives, the analyst chooses the alternative (if there effective fighter aircraft - no Hurricanes or Spitfires had come into service - and no
is one) that best meets the organization's objectives. radar data was yet fed into its very elementary warning and control system.
Sometimes the set of alternatives is subject to certain restrictions and constraints. In It had become clear that radar would create a whole new series of problems in fighter
many situations, the best alternative may be impossible or too costly to determine. direction and control so in late 1936 some experiments started at Biggin Hill in Kent
Step 6. Present the Results and Conclusions of the Study into the effective use of such data. This early work, attempting to integrate radar data
In this step, the analyst presents the model and the recommendations from Step 5 to with ground based observer data for fighter interception, was the start of OR.
the decision making individual or group. In some situations, one might present The first of three major pre-war air-defense exercises was carried out in the summer
several alternatives and let the organization choose the decision maker(s) choose of 1937. The experimental radar station at Bawdsey Research Station was brought
the one that best meets her/his/their needs. into operation and the information derived from it was fed into the general air-defense
After presenting the results of the OR study to the decision maker(s), the analyst may warning and control system. From the early warning point of view this exercise was
find that s/he does not (or they do not) approve of the recommendations. This may encouraging, but the tracking information obtained from radar, after filtering and
result from incorrect definition of the problem on hand or from failure to involve transmission through the control and display network, was not very satisfactory.
decision maker(s) from the start of the project. In this case, the analyst should return In July 1938 a second major air-defense exercise was carried out. Four additional
to Step 1, 2, or 3. radar stations had been installed along the coast and it was hoped that Britain now
Step 7. Implement and Evaluate Recommendation had an aircraft location and control system greatly improved both in coverage and
If the decision maker(s) has accepted the study, the analyst aids in implementing the effectiveness. Not so! The exercise revealed, rather, that a new and serious problem
recommendations. The system must be constantly monitored (and updated had arisen. This was the need to coordinate and correlate the additional, and often
dynamically as the environment changes) to ensure that the recommendations are conflicting, information received from the additional radar stations. With the out-break
enabling decision maker(s) to meet her/his/their objectives. of war apparently imminent, it was obvious that something new - drastic if necessary
- had to be attempted. Some new approach was needed.
1.3 HISTORY OF OR Accordingly, on the termination of the exercise, the Superintendent of Bawdsey
(Prof. Beasley’s lecture notes) Research Station, A.P. Rowe, announced that although the exercise had again
OR is a relatively new discipline. Whereas 70 years ago it would have been possible demonstrated the technical feasibility of the radar system for detecting aircraft, its
to study mathematics, physics or engineering (for example) at university it would not operational achievements still fell far short of requirements. He therefore proposed
have been possible to study OR, indeed the term OR did not exist then. It was only that a crash program of research into the operational - as opposed to the technical -
really in the late 1930's that operational research began in a systematic fashion, and aspects of the system should begin immediately. The term "operational research"
it started in the UK. [RESEARCH into (military) OPERATIONS] was coined as a suitable description of
Early in 1936 the British Air Ministry established Bawdsey Research Station, on the this new branch of applied science. The first team was selected from amongst the
east coast, near Felixstowe, Suffolk, as the centre where all pre-war radar scientists of the radar research group the same day.

2 3
In the summer of 1939 Britain held what was to be its last pre-war air defense OR started just before World War II in Britain with the establishment of teams of
exercise. It involved some 33,000 men, 1,300 aircraft, 110 antiaircraft guns, 700 scientists to study the strategic and tactical problems involved in military operations.
searchlights, and 100 barrage balloons. This exercise showed a great improvement The objective was to find the most effective utilization of limited military resources by
in the operation of the air defense warning and control system. The contribution the use of quantitative techniques.
made by the OR teams was so apparent that the Air Officer Commander-in-Chief Following the end of the war OR spread, although it spread in different ways in the
RAF Fighter Command (Air Chief Marshal Sir Hugh Dowding) requested that, on the UK and USA.
outbreak of war, they should be attached to his headquarters at Stanmore. You should be clear that the growth of OR since it began (and especially in the last
On May 15th 1940, with German forces advancing rapidly in France, Stanmore 30 years) is, to a large extent, the result of the increasing power and widespread
Research Section was asked to analyze a French request for ten additional fighter availability of computers. Most (though not all) OR involves carrying out a large
squadrons (12 aircraft a squadron) when losses were running at some three number of numeric calculations. Without computers this would simply not be
squadrons every two days. They prepared graphs for Winston Churchill (the British possible.
Prime Minister of the time), based upon a study of current daily losses and
replacement rates, indicating how rapidly such a move would deplete fighter strength.
No aircraft were sent and most of those currently in France were recalled.
This is held by some to be the most strategic contribution to the course of the war
made by OR (as the aircraft and pilots saved were consequently available for the
successful air defense of Britain, the Battle of Britain).
In 1941 an Operational Research Section (ORS) was established in Coastal
Command which was to carry out some of the most well-known OR work in World
War II.
Although scientists had (plainly) been involved in the hardware side of warfare
(designing better planes, bombs, tanks, etc) scientific analysis of the operational use
of military resources had never taken place in a systematic fashion before the
Second World War. Military personnel, often by no means stupid, were simply not
trained to undertake such analysis.
These early OR workers came from many different disciplines, one group consisted
of a physicist, two physiologists, two mathematical physicists and a surveyor. What
such people brought to their work were "scientifically trained" minds, used to querying
assumptions, logic, exploring hypotheses, devising experiments, collecting data,
analyzing numbers, etc. Many too were of high intellectual caliber (at least four
wartime OR personnel were later to win Nobel prizes when they returned to their
peacetime disciplines).
By the end of the war OR was well established in the armed services both in the UK
and in the USA.

4 5
Mine Cost per day (£'000) Production (tons/day)
High Medium Low
X 180 6 3 4
2. BASIC OR CONCEPTS Y 160 1 1 6
How many days per week should each mine be operated to fulfill the smelting plant
"OR is the representation of real-world systems by mathematical models together contract?
with the use of quantitative methods (algorithms) for solving such models, with a view
to optimizing." Guessing
We can also define a mathematical model as consisting of: To explore the Two Mines problem further we might simply guess (i.e. use our
! Decision variables, which are the unknowns to be determined by the solution to judgment) how many days per week to work and see how they turn out.
the model. ! work one day a week on X, one day a week on Y
! Constraints to represent the physical limitations of the system This does not seem like a good guess as it results in only 7 tones a day of high-
! An objective function grade, insufficient to meet the contract requirement for 12 tones of high-grade a day.
! An optimal solution to the model is the identification of a set of variable values We say that such a solution is infeasible.
which are feasible (satisfy all the constraints) and which lead to the optimal value
! work 4 days a week on X, 3 days a week on Y
of the objective function.
This seems like a better guess as it results in sufficient ore to meet the contract. We
In general terms we can regard OR as being the application of scientific methods /
say that such a solution is feasible. However it is quite expensive (costly).
thinking to decision making.
We would like a solution which supplies what is necessary under the contract at
Underlying OR is the philosophy that:
minimum cost. Logically such a minimum cost solution to this decision problem must
! decisions have to be made; and
exist. However even if we keep guessing we can never be sure whether we have
! using a quantitative (explicit, articulated) approach will lead to better decisions
found this minimum cost solution or not. Fortunately our structured approach will
than using non-quantitative (implicit, unarticulated) approaches.
enable us to find the minimum cost solution.
Indeed it can be argued that although OR is imperfect it offers the best available
approach to making a particular decision in many instances (which is not to say that
Solution
using OR will produce the right decision).
What we have is a verbal description of the Two Mines problem. What we need to do
is to translate that verbal description into an equivalent mathematical description.
Two Mines Example
In dealing with problems of this kind we often do best to consider them in the order:
The Two Mines Company own two different mines that produce an ore which, after
! Variables
being crushed, is graded into three classes: high, medium and low-grade. The
! Constraints
company has contracted to provide a smelting plant with 12 tons of high-grade, 8
! Objective
tons of medium-grade and 24 tons of low-grade ore per week. The two mines have
This process is often called formulating the problem (or more strictly formulating a
different operating characteristics as detailed below.
mathematical representation of the problem).

6 7
Variables Implicit constraints
These represent the "decisions that have to be made" or the "unknowns". Constraints such as days per week constraint are often called implicit constraints
Let because they are implicit in the definition of the variables.
x = number of days per week mine X is operated
y = number of days per week mine Y is operated Objective
Note here that x >= 0 and y >= 0. Again in words our objective is (presumably) to minimize cost which is given by 180x
+ 160y
Constraints
It is best to first put each constraint into words and then express it in a mathematical Hence we have the complete mathematical representation of the problem:
form. minimize
180x + 160y
ore production constraints - balance the amount produced with the
subject to
quantity required under the smelting plant contract 6x + y >= 12
3x + y >= 8
Ore
4x + 6y >= 24
High 6x + 1y >= 12 x <= 5
y <= 5
Medium 3x + 1y >= 8
x,y >= 0
Low 4x + 6y >= 24
days per week constraint - we cannot work more than a certain Some notes
maximum number of days a week e.g. for a 5 day week we have The mathematical problem given above has the form
x <= 5 ! all variables continuous (i.e. can take fractional values)
y <= 5 ! a single objective (maximize or minimize)
! the objective and constraints are linear i.e. any term is either a constant or a
Inequality constraints constant multiplied by an unknown (e.g. 24, 4x, 6y are linear terms but xy is a
Note we have an inequality here rather than an equality. This implies that we may non-linear term)
produce more of some grade of ore than we need. In fact we have the general rule: Any formulation which satisfies these three conditions is called a linear program (LP).
given a choice between an equality and an inequality choose the inequality We have (implicitly) assumed that it is permissible to work in fractions of days -
For example - if we choose an equality for the ore production constraints we have the problems where this is not permissible and variables must take integer values will be
three equations 6x+y=12, 3x+y=8 and 4x+6y=24 and there are no values of x and y dealt with under Integer Programming (IP).
which satisfy all three equations (the problem is therefore said to be "over-
constrained"). For example the values of x and y which satisfy 6x+y=12 and 3x+y=8
are x=4/3 and y=4, but these values do not satisfy 4x+6y=24.
The reason for this general rule is that choosing an inequality rather than an equality
gives us more flexibility in optimizing (maximizing or minimizing) the objective
(deciding values for the decision variables that optimize the objective).

8 9
Discussion
This problem was a decision problem.
We have taken a real-world situation and constructed an equivalent mathematical 3. LINEAR PROGRAMMING
representation - such a representation is often called a mathematical model of the
real-world situation (and the process by which the model is obtained is called It can be recalled from the Two Mines example that the conditions for a mathematical
formulating the model). model to be a linear program (LP) were:
Just to confuse things the mathematical model of the problem is sometimes called ! all variables continuous (i.e. can take fractional values)
the formulation of the problem. ! a single objective (minimize or maximize)

Having obtained our mathematical model we (hopefully) have some quantitative ! the objective and constraints are linear i.e. any term is either a constant or a
method which will enable us to numerically solve the model (i.e. obtain a numerical constant multiplied by an unknown.
solution) - such a quantitative method is often called an algorithm for solving the LP's are important - this is because:
model. ! many practical problems can be formulated as LP's
Essentially an algorithm (for a particular model) is a set of instructions which, when ! there exists an algorithm (called the simplex algorithm) which enables us to solve
followed in a step-by-step fashion, will produce a numerical solution to that model. LP's numerically relatively easily
Our model has an objective, that is something which we are trying to optimize. We will return later to the simplex algorithm for solving LP's but for the moment we
Having obtained the numerical solution of our model we have to translate that will concentrate upon formulating LP's.
solution back into the real-world situation. Some of the major application areas to which LP can be applied are:
! Work scheduling
! Production planning & Production process
"OR is the representation of real-world systems by mathematical models ! Capital budgeting
together with the use of quantitative methods (algorithms) for solving such ! Financial planning
models, with a view to optimizing." ! Blending (e.g. Oil refinery management)
! Farm planning
! Distribution
! Multi-period decision problems
o Inventory model
o Financial models
o Work scheduling
Note that the key to formulating LP's is practice. However a useful hint is that
common objectives for LP's are maximize profit/minimize cost.

10 11
There are four basic assumptions in LP:
3.1 FORMULATING LP
! Proportionality
o The contribution to the objective function from each decision variable is 3.1.1 Giapetto Example
proportional to the value of the decision variable (The contribution to the (Winston 3.1, p. 49)
objective function from making four soldiers (4"$3=$12) is exactly four Giapetto's wooden soldiers and trains. Each soldier sells for $27, uses $10 of raw
times the contribution to the objective function from making one soldier materials and takes $14 of labor & overhead costs. Each train sells for $21, uses $9
($3)) of raw materials, and takes $10 of overhead costs. Each soldier needs 2 hours
o The contribution of each decision variable to the LHS of each constraint is finishing and 1 hour carpentry; each train needs 1 hour finishing and 1 hour
proportional to the value of the decision variable (It takes exactly three carpentry. Raw materials are unlimited, but only 100 hours of finishing and 80 hours
times as many finishing hours (2hrs"3=6hrs) to manufacture three soldiers of carpentry are available each week. Demand for trains is unlimited; but at most 40
soldiers can be sold each week. How many of each toy should be made each week
as it takes to manufacture one soldier (2 hrs))
to maximize profits.
! Additivity
o The contribution to the objective function for any decision variable is
Answer
independent of the values of the other decision variables (No matter what
Decision variables completely describe the decisions to be made (in this case, by
the value of train (x2), the manufacture of soldier (x1) will always contribute
Giapetto). Giapetto must decide how many soldiers and trains should be
3x1 dollars to the objective function)
manufactured each week. With this in mind, we define:
o The contribution of a decision variable to LHS of each constraint is
x1 = the number of soldiers produced per week,
independent of the values of other decision variables (No matter what the
x2 = the number of trains produced per week,
value of x1, the manufacture of x2 uses x2 finishing hours and x2 carpentry
Objective function is the function of the decision variables that the decision maker
hours)
wants to maximize (revenue or profit) or minimize (costs). Giapetto can concentrate
1st implication: The value of objective function is the sum of the contributions
on maximizing the total weekly profit (z).
from each decision variables.
Here profit equals to (weekly revenues) – (raw material purchase cost) – (other
2nd implication: LHS of each constraint is the sum of the contributions from
variable costs). Hence Giapetto’s objective function is:
each decision variables.
Maximize z = 3x1 + 2x2
! Divisibility
Constraints show the restrictions on the values of the decision variables. Without
o Each decision variable is allowed to assume fractional values. If we
constraints Giapetto could make a large profit by choosing decision variables to be
actually can not produce a fractional number of decision variables, we use
very large. Here there are three constraints:
IP (It is acceptable to produce 1.69 trains)
Finishing time per week
! Certainty
Carpentry time per week
o Each parameter is known with certainty
Weekly demand for soldiers
Sign restrictions are added if the decision variables can only assume nonnegative
values (Giapetto can not manufacture negative number of soldiers or trains!)

12 13
All these characteristics explored above give the following Linear Programming 3.1.2 Advertisement Example
(LP) problem (Winston 3.2, p.61)
max z = 3x1 + 2x2 (The Objective function) Dorian makes luxury cars and jeeps for high-income men and women. It wishes to
s.t. 2x1 + x2 # 100 (Finishing constraint) advertise with 1 minute spots in comedy shows and football games. Each comedy
x1 + x2! # 80 (Carpentry constraint) spot costs $50K and is seen by 7M high-income women and 2M high-income men.
x1 # 40 (Constraint on demand for soldiers) Each football spot costs $100K and is seen by 2M high-income women and 12M
x1, x2 > 0 (Sign restrictions) high-income men. How can Dorian reach 28M high-income women and 24M high-
A value of (x1,x2) is in the feasible region if it satisfies all the constraints and sign income men at the least cost.
restrictions.
Graphically and computationally we see the solution is (x1,x2) = (20,60) at which z = Answer: The decision variables are
180. (Optimal solution) x1 = the number of comedy spots

x2 = the number of football spots.
Report
Giving the problem
The maximum profit is $180 by making 20 soldiers and 60 trains each week. Profit is
limited by the carpentry and finishing labor available. Profit could be increased by min z = 50x1 + 100x2
buying more labor. st 7x1 + 2x2 $% 28

2x1 + 12x2 %$%% 24
x1, x2 $% 0
The graphical solution is z = 320 when (x1,x2) = (3.6,1.4). From the graph, in this
problem rounding up to (x1,x2) = (4,2) gives the best integer solution.
Report: The minimum cost of reaching the target audience is $400K, with 4 comedy
spots and 2 football slots. The model is dubious as it does not allow for saturation
after repeated viewings.

14 15
3.1.3 Diet Example 3.1.4 Post Office Example
(Winston 3.4., p. 70) (Winston 3.5, p.74)
Ms. Fidan’s diet requires that all the food she eats come from one of the four “basic A PO requires different numbers of employees on different days of the week. Union
food groups“. At present, the following four foods are available for consumption: rules state each employee must work 5 consecutive days and then receive two days
brownies, chocolate ice cream, cola, and pineapple cheesecake. Each brownie costs off. Find the minimum number of employees needed.
0.5$, each scoop of chocolate ice cream costs 0.2$, each bottle of cola costs 0.3$,
Mon Tue Wed Thur Fri Sat Sun
and each pineapple cheesecake costs 0.8$. Each day, she must ingest at least 500
Staff Needed 17 13 15 19 14 16 11
calories, 6 oz of chocolate, 10 oz of sugar, and 8 oz of fat. The nutritional content per
unit of each food is shown in Table. Formulate an LP model that can be used to Answer: The decision variables are xi (# of employees starting on day i)
satisfy her daily nutritional requirements at minimum cost. Mathematically we must
Calories Chocolate Sugar Fat min z = x1 +x2 +x3 +x4 +x5 +x6 +x7
(ounces) (ounces) (ounces)
Brownie 400 3 2 2 x1 +x4 +x5 +x6 +x7 %$% 17
Choc. ice cream (1 scoop) 200 2 2 4 x1 +x2 +x5 +x6 +x7 %$% 13
Cola (1 bottle) 150 0 4 1
Pineapple cheesecake (1 piece) 500 0 4 5 x1 +x2 +x3 +x6 +x7 %$% 15
x1 +x2 +x3 +x4 +x7 %$% 19
Answer
x1 +x2 +x3 +x4 +x5 %$% 14
The decision variables:
+x2 +x3 +x4 +x5 +x6 %$% 16
x1: number of brownies eaten daily
+x3 +x4 +x5 +x6 +x7 %$% 11
x2: number of scoops of chocolate ice cream eaten daily
x3: bottles of cola drunk daily
The solution is (xi) = (4/3,10/3,2,22/3,0,10/3,5) giving z = 67/3. We could round this
x4: pieces of pineapple cheesecake eaten daily
up to (xi) = (2,4,2,8,0,4,5) giving z = 25. However restricting the decision var.s to be
The objective function (the total cost of the diet in cents):
integers and using Lindo again gives (xi) = (4,4,2,6,0,4,3) giving z = 23
min w = 50 x1 + 20 x2 + 30 x3 + 80 x4
Constraints:
400 x1 + 200 x2 + 150 x3 + 500 x4 > 500 (daily calorie intake)
3 x1 + 2 x2 > 6 (daily chocolate intake)
2 x1 + 2 x2 + 4 x3 + 4 x4 > 10 (daily sugar intake)
2 x1 + 4 x2 + x3 + 5 x4 > 8 (daily fat intake)
xi > 0, i = 1, 2, 3, 4 (Sign restrictions!)
Report:
The minimum cost diet incurs a daily cost of 90 cents by eating 3 scoops of chocolate and
drinking 1 bottle of cola (w=90, x2=3, x3=1)

16 17
3.1.5 Sailco Example 3.1.6 Customer Service Level Example
(Winston 3.10, p. 99) (Winston 3.12, p. 108)
Sailco must determine how many sailboats to produce in the next 4 quarters The CSL services computers. Its demand (hours) for the time of skilled technicians in the
demand is known to be 40, 60, 75, and 25 boats. Sailco must meet its demands. At next 5 months is
st
the beginning of the 1 Sailco starts with 10 boats in inventory. Sailco can produce t Jan Feb Mar Apr May
up to 40 boats with regular time labor at $400 per boat, or additional boats at $450 dt 6000 7000 8000 9500 11000
with overtime labor. Boats made in a quarter can be used to meet that quarter's It starts with 50 skilled technicians at the beginning of January. Each technician can
demand or held in inventory for the next quarter at an extra cost of $20.00 per boat. work 160 hrs/month. To train a new technician they must be supervised for 50 hrs by
an experienced technician. Each experienced technician is paid $2K/mth and a
Answer: trainee is paid $1K/mth. Each month 5% of the skilled technicians leave. CSL needs
The decision variables are for t = 1,2,3,4 to meet demand and minimize costs.
xt = # of boats in quarter t built in regular time
yt = # of boats in quarter t built in overtime Answer:
For convenience, introduce The decision variable is
it = # of boats in inventory at the end quarter t xt = # to be trained in month t
dt = demand in quarter t We must minimize the total cost. For convenience let
We are given that xt < 40, &t yt = # experienced tech. at start of t th month
By logic it = it-1+ xt + yt - dt, &t. dt = demand during month t
Demand is met iff it > 0, &t Then we must
(Sign restrictions xt and yt > 0, &t) min z = 2000(y1+...+y5)+1000(x1+...+x5)
We need to minimize total cost z subject to these three sets of conditions where subject to
z = 400(x1+x2+x3+x4) + 450(y1+y2+y3+y4) + 20(i1+i2+i3+i4)
160yt-50xt > dt for t = 1,...5
y1 = 50
Report:
Lindo reveals the solution to be (x1, x2, x3, x4) = (40, 40, 40, 25) and (y1, y2, y3, y4) = yt = .95yt-1+xt-1 for t = 2,3,4,5
(0, 10, 35, 0) and the minimum cost of $78450.00 is achieved by the schedule
xt, yt > 0
Q1 Q2 Q3 Q4
Regular time (xt) 40 40 40 25
Overtime (yt) 0 10 35 0
Inventory (it) 10 10 0 0 0
Demand (dt) 40 60 75 25

18 19
3.2 SOLVING LP 3.2.2 The Graphical Solution (Minimization)
3.2.1 LP Solutions: Four Cases min 180x + 160y
st 6x + y >= 12
When an LP is solved, one of the following four cases will occur:
3x + y >= 8
1. The LP has a unique optimal solution. 4x + 6y >= 24
x <= 5
2. The LP has alternative (multiple) optimal solutions. It has more than one
y <= 5
(actually an infinite number of) optimal solutions x,y >= 0
3. The LP is infeasible. It has no feasible solutions (The feasible region contains Since there are only two variables in this LP problem we have the graphical
no points). representation of the LP given below with the feasible region (region of feasible
4. The LP is unbounded. In the feasible region there are points with arbitrarily solutions to the constraints associated with the LP) outlined.
large (in a max problem) objective function values.
Graphically,
! In the unique optimal solution case, isoprofit line last hits a point (vertex -
corner) before leaving the feasible region.
! In the alternative optimal solutions case, isoprofit line last hits an entire line
(representing one of the constraints) before leaving the feasible region.
! In the infeasible case, there is no feasible region.
! In the unbounded case, isoprofit line never lose contact with the feasible
region
We determine the optimal solution to the LP by plotting 180x + 160y = K (K constant)

for varying K values (isoprofit lines). One such line (180x + 160y = 180) is shown
dotted on the diagram. The smallest value of K (remember we are considering a
minimization problem) such that 180x + 160y = K goes through a point in the feasible
region is the value of the optimal solution to the LP (and the corresponding point
gives the optimal values of the variables). Hence we can see that the optimal solution
to the LP occurs at the vertex of the feasible region formed by the intersection of 3x +
y = 8 and 4x + 6y = 24.
Optimal sol’n is 765.71 (1.71 days mine x and 2.86 days mine y are operated)

20 21
3.2.3 The Graphical Solution (Maximization) 3.2.4 The Simplex Algorithm
Max z = 4x1 + 2x2 (The Objective function) Note that in the example considered at the graphical solution (min) section, the
s.t. 2x1 + x2 < 100 (Finishing constraint) optimal solution to the LP occurred at a vertex (corner) of the feasible region. In fact it
x1 + x2< 80 (Carpentry constraint) is true that for any LP the optimal solution occurs at a vertex of the feasible region.
x1 < 40 (Constraint on demand for soldiers) This fact is the key to the simplex algorithm for solving LP's.
x1 + x2 > 0 (Sign restrictions) Essentially the simplex algorithm starts at one vertex of the feasible region and
moves (at each iteration) to another (adjacent) vertex, improving (or leaving
x2
unchanged) the objective function as it does so, until it reaches the vertex
100 B corresponding to the optimal LP solution.
The simplex algorithm for solving linear programs (LP's) was developed by Dantzig in
80 D the late 1940's and since then a number of different versions of the algorithm have
been developed. One of these later versions, called the revised simplex algorithm
G (sometimes known as the "product form of the inverse" simplex algorithm) forms the
basis of most modern computer packages for solving LP's.
Steps
1. Convert the LP to standard form
2. Obtain a basic feasible solution (bfs) from the standard form
F 3. Determine whether the current bfs is optimal. If it is optimal, stop.
4. If the current bfs is not optimal, determine which nonbasic variable should
E A C become a basic variable and which basic variable should become a nonbasic
x1 variable to find a new bfs with a better objective function value
H 40 50 80
5. Go back to Step 3.
Points on the line between points G (20, 60) and F (40, 20) are the alternative
optimal solutions. Thus, for 0<c<1,
Related concepts:
c [20 60] + (1-c) [40 20] = [40-20c 20+40c]
! Standard form: all constraints are equations and all variables are nonnegative
will be optimal
! bfs: any basic solution where all variables are nonnegative
********************************************************************* ! Nonbasic variable: a chosen set of variables where variables equal to 0
Add constraint x2 > 90 (Constraint on demand for trains)
! Basic variable: the remaining variables that satisfy the system of equations at
No feasible region: Infeasible LP
the standard form
*********************************************************************
Only use constraint x2 > 90
Isoprofit line never lose contact with the feasible region: Unbounded LP

22 23
3.2.5 Example illustrating Simplex (Dakota Furniture) Solution with Simplex Algorithm
(Winston 4.3, p. 134) 1. First introduce slack variables and convert the LP to the standard form and write a
Dakota Furniture makes desks, tables, and chairs. Each product needs the limited canonical form
resources of lumber, carpentry and finishing; as described in the table. At most 5
tables can be sold per week. Maximize weekly revenue. max z = 60x1 + 30x2 + 20x3
s.t. 8x1 + 6x2 + x3 +s1 = 48
Resource Desk Table Chair Max Avail.
4x1 + 2x2 + 1.5x3 +s2 = 20
Lumber (board ft.) 8 6 1 48
2x1 + 1.5x2 + .5x3 +s3 = 8
Finishing hours 4 2 1.5 20
x2 +s4 = 5
Carpentry hours 2 1.5 .5 8
x1,x2,x3,s1,s2,s3,s4 $ 0.
Max Demand unlimited 5 unlimited
Price ($) 60 30 20
R0 z -60x1 -30x2 -20x3 =0
R1 8x1 + 6x2 + x3 +s1 = 48
LP Model: R2 4x1 + 2x2 +1.5x3 +s2 = 20
Let x1, x2, x3 be the number of desks, tables and chairs produced. Let the weekly R3 2x1 +1.5x2 + .5x3 +s3 =8
profit be $z. Then, we must R4 x2 +s4 = 5
max z = 60x1+30x2+20x3
s.t. 8x1+ 6x2+ x3 < 48 2. Obtain a starting bfs.
4x1+ 2x2+1.5x3 < 20 As (x1, x2, x3) = 0 is feasible for the original problem, the below given point where
2x1+1.5x2+. 5x3 < 8 three of the variables equal 0 (the non-basic variables) and the four other variables
x2 < 5 (the basic variables) are determined by the four equalities is an obvious bfs:
x1,x2,x3 > 0 x1 = x2 = x3 = 0, s1 = 48, s2 = 20, s3 = 8, s4 = 5.
.
3. Determine whether the current bfs is optimal.
Determine whether there is any way that z can be increased by increasing some
nonbasic variable.
If each nonbasic variable has a nonnegative coefficient in the objective function row
(row 0), current bfs is optimal.
However, here all nonbasic variables have negative coefficients: It is not optimal.
4. Find a new bfs

! z increases most rapidly when x1 is made non-zero; i.e. x1 is the entering
variable.

24 25
! Examining R1, x1 can be increased only to 6. More than 6 makes s1 < 0.
R0’’ z +5x2 + 10s2+10s3 = 280 z = 280 (R0’+5R2’’)
Similarly R2, R3, and R4, give limits of 5, 4, and no limit for x1 (ratio test). The
smallest ratio is the largest value of the entering variable that will keep all the R1’’ -2x2 + s1 2s2- 8s3 = 24 s1 = 24 (R1’+R2’’)
current basic variables nonnegative. Thus by R3, x1 can only increase to x1 = 4 R2’’ -2x2 +x3 + 2s2- 4s3 =8 x3 = 8 (R2’’)
when s3 becomes 0. (I.e. look for the least positive of the ratios 48/8 = 6, 20/4
R3’’ x1+1.25x2 - .5s2+1.5s3 =2 x1 = 2 (R3’-.5R2’’)
= 5, 8/2 = 4, 5/0 = '). We say s3 is the leaving variable and R3 is the pivot
equation. R4’’ x2 + s4 = 5 s4 = 5 (R4’)
! Now we must rewrite the system so the values of the basic variables can be The bfs is now x2 = s2 = s3 = 0, x1 = 2, x3 = 8, s1 = 24, s4 = 5 making z = 280.
read off.
Each nonbasic variable has a nonnegative coefficient in row 0
The new pivot equation (R3/2) is THE CURRENT SOLUTION IS OPTIMAL
R3’ : x1+.75x2+.25x3+ .5s3 =4

Report: Dakota furniture’s optimum weekly profit would be 280$ if they produce 2
’
Then use R3 to eliminate x1 in all the other rows. desks and 8 chairs.
R0’ z +15x2 -5x3 +30s3 = 240 z = 240 (R0+60R3’

R1’ -x3 +s1 -4s3 = 16 s1 = 16 (R1-8R3’)
’
R2 -x2 +.5x3 + s2-2s3 =4 s2 = 4 (R2-4R3’)
R3’ x1+.75x2 +.25x3 +.5s3 =4 x1 = 4 (R3’)
R4’ x2 + s4 = 5 s4 = 5 (R4)
The new bfs is x2 = x3 = s3 = 0, x1 = 4, s1 = 16, s2 = 4, s4 = 5 making z = 240.
5. Go back to Step 3. Repeat steps until an optimal solution is reached

! We increase z fastest by making x3 non-zero (i.e. x3 enters).
! x3 can be increased to at most x3 = 8, when s2 = 0 ( i.e. s2 leaves.)
Rearranging the pivot equation gives
R2’’ -2x2+x3+2s2-4s3 = 8 (R2’×2).
Row operations with R2’’ eliminate x3 to give the new system

26 27
This was once written as a tableau. 3.2.6 Other Solution Cases by Simplex
(Use tableau format for each operation in all HW and exams!!!)
Alternative Optimal Solutions
Initial tableau: Dakota example is modified: $35/table
z x1 x2 x3 s1 s2 s3 s4 RHS BV new z = 60 x1 + 35 x2 + 20 x3
1 -60 -30 -20 0 0 0 0 0 z=0 Second and optimal tableau for the modified problem:
(
8 6 1 1 0 0 0 48 s1 = 48 z x1 x2 x3 s1 s2 s3 s4 RHS BV Ratio
4 2 1.5 0 1 0 0 20 s2 = 20 1 0 0 0 0 10 10 0 280 z=280
2 1.5 .5 0 0 1 0 8 s3 = 8 0 0 -2 0 1 2 -8 0 24 s1=24 -
0 1 0 0 0 0 1 5 s4 = 5 0 0 -2 1 0 2 -4 0 8 x3=8 -
0 1 1.25 0 0 -.5 1.5 0 2 x1=2 2/1.25 )

First tableau:
0 0 1 0 0 0 0 1 5 s4=5 5/1
z x1 x2 x3 s1 s2 s3 s4 RHS BV
Another optimal tableau for the modified problem:
1 15 -5 30 240 z = 240
-1 1 -4 16 s1 = 16
1 0 0 0 0 10 10 0 280 z=280
-1 .5 1 -2 4 s2 = 4
0 1.6 0 0 1 1.2 -5.6 0 27.2 s1=27.2
1 .75 .25 .5 4 x1 = 4
0 1.6 0 1 0 1.2 -1.6 0 11.2 x3=11.2
1 1 5 s4 = 5
0 0.8 1 0 0 -0.4 1.2 0 1.6 x2=1.6
Second and optimal tableau: 0 -0.8 0 0 0 0.4 -1.2 1 3.4 s4=3.4
Therefore the optimal solution is as follows:
1 0 5 0 0 10 10 0 280 z = 280 z = 280 and for 0 < c < 1
0 -2 0 1 2 -8 0 24 s1 = 24
x1 2 0 2c
0 -2 1 0 2 -4 0 8 x3 = 8 x2 = c 0 + (1–c) 1.6 = 1.6 – 1.6c
1 1.25 0 0 -.5 1.5 0 2 x1 = 2 x3 8 11.2 11.2 – 3.2c
0 1 0 0 0 0 1 5 s4 = 5

28 29
Unbounded LPs 3.2.7 The Big M Method
If an LP has any > or = constraints, a starting bfs may not be readily apparent.
(
z x1 When a bfs is not readily apparent, the Big M method or the two-phase simplex
x2 x3 x4 s1 s2 RHS BV Ratio
method may be used to solve the problem.
1 0 2 -9 0 12 4 100 z=100
The Big M method is a version of the Simplex Algorithm that first finds a bfs by
0 0 1 -6 1 6 -1 20 x4=20 None adding "artificial" variables to the problem. The objective function of the original LP
0 1 1 -1 0 1 0 5 x1=5 None must, of course, be modified to ensure that the artificial variables are all equal to 0 at
the conclusion of the simplex algorithm.
Since ratio test fails, the LP under consideration is an unbounded LP.
Steps
1. Modify the constraints so that the RHS of each constraint is nonnegative (This
requires that each constraint with a negative RHS be multiplied by -1. Remember
that if you multiply an inequality by any negative number, the direction of the
inequality is reversed!). After modification, identify each constraint as a <, >, or =
constraint.
2. Convert each inequality constraint to standard form (If constraint i is a <
constraint, we add a slack variable si; and if constraint i is a > constraint, we
subtract an excess variable ei).
3. Add an artificial variable ai to the constraints identified as > or = constraints at the
end of Step 1. Also add the sign restriction ai > 0.
4. Let M denote a very large positive number. If the LP is a min problem, add (for
each artificial variable) Mai to the objective function. If the LP is a max problem,
add (for each artificial variable) -Mai to the objective function.
5. Since each artificial variable will be in the starting basis, all artificial variables
must be eliminated from row 0 before beginning the simplex. Now solve the
transformed problem by the simplex (In choosing the entering variable, remember
that M is a very large positive number!).
If all artificial variables are equal to zero in the optimal solution, we have found the
optimal solution to the original problem.
If any artificial variables are positive in the optimal solution, the original problem is
infeasible!!!

30 31
3.2.8 Example illustrating Big M (Oranj Juice) 3. Add ai to the constraints identified as > or = const.s
(Winston 4.10, p. 164) z– 2 x1 – 3 x2 = 0 Row 0
Bevco manufactures an orange flavored soft drink called Oranj by combining orange 0.5 x1+ 0.25 x2 + s1 = 4 Row 1
soda and orange juice. Each ounce of orange soda contains 0.5 oz of sugar and 1 x1+ 3 x2 - e2 + a2 = 20 Row 2
mg of vitamin C. Each ounce of orange juice contains 0.25 oz of sugar and 3 mg of x1+ x2 + a3 = 10 Row 3
vitamin C. It costs Bevco 2¢ to produce an ounce of orange soda and 3¢ to produce all variables nonnegative
an ounce of orange juice. Marketing department has decided that each 10 oz bottle
of Oranj must contain at least 20 mg of vitamin C and at most 4 oz of sugar. Use LP 4. Add Mai to the objective function (min problem)
to determine how Bevco can meet marketing dept.’s requirements at minimum cost. min z = 2 x1 + 3 x2 + M a2 + M a3
Row 0 will change to
LP Model: z– 2 x1 – 3 x2 – M a2 – M a3 = 0
Let x1 and x2 be the quantity of ounces of orange soda and orange juice
(respectively) in a bottle of Oranj. 5. Since each artificial variable are in our starting bfs, they must be eliminated from
min z = 2x1+3x2 row 0
s.t. 0.5 x1+ 0.25 x2 < 4 (sugar const.) New Row 0 = Row 0 + M * Row 2 + M * Row 3 )
x1+ 3 x2 > 20 (vit. C const.) z + (2M–2) x1 + (4M–3) x2 – M e2 = 30M New Row 0
x1+ x2 = 10 (10 oz in bottle)
x1,x2 > 0 Initial tableau:
(
z x1 x2 s1 e2 a2 a3 RHS BV Ratio
1 2M-2 4M-3 0 -M 0 0 30M z=30M
Solving Oranj Example with Big M Method 0 0.5 0.25 1 0 0 0 4 s1=4 16
0 1 3 0 -1 1 0 20 a2=20 20/3*
0 1 1 0 0 0 1 10 a3=10 10
1. Modify the constraints so that the RHS of each constraint is nonnegative
The RHS of each constraint is nonnegative In a min problem, entering variable is the variable that has the “most positive”
coefficient in row 0!
2. Convert each inequality constraint to standard form

z– 2 x1 – 3 x2 = 0 First tableau:
0.5 x1+ 0.25 x2 + s1 = 4 (

z x1 x2 s1 e2 a2 a3 RHS BV Ratio
x1+ 3 x2 - e2 = 20 1 (2M-3)/3 0 0 (M-3)/3 (3-4M)/3 0 20+3.3M z
x1+ x2 = 10 0 5/12 0 1 1/12 -1/12 0 7/3 s1 28/5
0 1/3 1 0 -1/3 1/3 0 20/3 x2 20
all variables nonnegative 0 2/3 0 0 1/3 -1/3 1 10/3 a3 5*

32 33
Optimal tableau:
3.3 SENSITIVITY ANALYSIS
z x1 x2 s1 e2 a2 a3 RHS BV
1 0 0 0 -1/2 (1-2M)/2 (3-2M)/2 25 z=25 3.3.1 Reduced Cost
0 0 0 1 -1/8 1/8 -5/8 1/4 s1=1/4
0 0 1 0 -1/2 1/2 -1/2 5 x2=5 For any nonbasic variable, the reduced cost for the variable is the amount by which
0 1 0 0 1/2 -1/2 3/2 5 x1=5 the nonbasic variable's objective function coefficient must be improved before that
variable will become a basic variable in some optimal solution to the LP.
Report:
If the objective function coefficient of a nonbasic variable xk is improved by its
In a bottle of Oranj, there should be 5 oz orange soda and 5 oz orange juice. In this
reduced cost, then the LP will have alternative optimal solutions -at least one in
case the cost would be 25¢.
which xk is a basic variable, and at least one in which xk is not a basic variable.
If the objective function coefficient of a nonbasic variable xk is improved by more than
its reduced cost, then any optimal solution to the LP will have xk as a basic variable
and xk>0.
Reduced cost of a basic variable is zero (see definition)!
3.3.2 Shadow Price

We define the shadow price for the ith constraint of an LP to be the amount by which
the optimal z value is "improved" (increased in a max problem and decreased in a
min problem) if the RHS of the ith constraint is increased by 1.
This definition applies only if the change in the RHS of the constraint leaves the
current basis optimal!
A > constraint will always have a nonpositive shadow price; a < constraint will always
have a nonnegative shadow price.

34 35
3.3.3 Simple Example 3.3.4 Utilizing Lindo Output for Sensitivity
NOTICE: The objective function which is regarded as Row 0 in Simplex is

max z = 5 x1 + x2 + 10 x3
accepted as Row 1 in Lindo.
x1 + x3 < 100
Therefore the first constraint of the model is always second row in Lindo!!!
x2 < 1
All variables > 0
MAX 5 X1 + X2 + 10 X3
SUBJECT TO
2) X1 + X3 <= 100
This is a very easy LP model and can be solved manually without utilizing Simplex.
3) X2 <= 1
x2 = 1 (This variable does not exist in the first constraint. In this case, as the problem END
is a maximization problem, the optimum value of the variable equals the RHS value
of the second constraint). LP OPTIMUM FOUND AT STEP 1
x1 = 0, x3 = 100 (These two variables do exist only in the first constraint and as the
OBJECTIVE FUNCTION VALUE
objective function coefficient of x3 is greater than that of x1, the optimum value of x3 1) 1001.000
equals the RHS value of the first constraint).
VARIABLE VALUE REDUCED COST
Hence, the optimal solution is as follows: X1 0.000000 5.000000
X2 1.000000 0.000000
z = 1001, [x1, x2, x3] = [0, 1, 100]
X3 100.000000 0.000000
ROW SLACK OR SURPLUS DUAL PRICES

Similarly, sensitivity analysis can be executed manually.
2) 0.000000 10.000000
Reduced Cost 3) 0.000000 1.000000
As x2 and x3 are in the basis, their reduced costs are 0.
RANGES IN WHICH THE BASIS IS UNCHANGED:
In order to have x1 enter in the basis, we should make its objective function
OBJ COEFFICIENT RANGES
coefficient as great as that of x3. In other words, improve the coefficient as 5 (10-5).
VARIABLE CURRENT ALLOWABLE ALLOWABLE
New objective function would be (max z = 10 x1 + x2 + 10 x3) and there would be at COEF INCREASE DECREASE
X1 5.000000 5.000000 INFINITY
least two optimal solutions for [x1, x2, x3]: [0, 1, 100] and [100, 1, 0].
X2 1.000000 INFINITY 1.000000
Therefore reduced cost of x1 equals 5. X3 10.000000 INFINITY 5.000000
If we improve the objective function coefficient of x1 more than its reduced cost, there
RIGHTHAND SIDE RANGES
would be a unique optimal solution: [100, 1, 0]. ROW CURRENT ALLOWABLE ALLOWABLE
RHS INCREASE DECREASE
Shadow Price
2 100.000000 INFINITY 100.000000
If the RHS of the first constraint is increased by 1, new optimal solution of x3 would 3 1.000000 INFINITY 1.000000
be 101 instead of 100. In this case, new z value would be 1011.
Lindo output reveals the reduced costs of x1, x2, and x3 as 5, 0, and 0 respectively.
If we use the definition: 1011 - 1001 = 10 is the shadow price of the fist constraint.
In the maximization problems, the reduced cost of a non-basic variable can also be
Similarly the shadow price of the second constraint can be calculated as 1 (please
read from the allowable increase value of that variable at obj. coefficient ranges.
find it).
Here, the corresponding value of x1 is 5.

36 37
In the minimization problems, the reduced cost of a non-basic variable can also be 3.3.5 Additional Example (Utilizing Simplex for Sensitivity)
read from the allowable decrease value of that variable at obj. coefficient ranges. In Dakota furniture example; x1, x2, and x3 were representing the number of desks,
tables, and chairs produced.
The same Lindo output reveals the shadow prices of the constraints in the "dual The LP formulated for profit maximization:
price" section:
max z = 60x1 + 30x2 + 20x3
Here, the shadow price of the first constraint (Row 2) equals 10.
The shadow price of the second constraint (Row 3) equals 1. 8x1 + 6x2 + x3 +s1 = 48 Lumber
4x1 + 2x2 + 1.5x3 +s2 = 20 Finishing
2x1 + 1.5x2 + .5x3 +s3 = 8 Carpentry

Some important equations
If the change in the RHS of the constraint leaves the current basis optimal (within the x2 +s4 = 5 Demand
allowable RHS range), the following equations can be used to calculate new
The optimal solution was:
objective function value:
for maximization problems z +5x2 + 10s2 + 10s3 = 280
! new obj. fn. value = old obj. fn. value + (new RHS – old RHS) × shadow price -2x2 +s1 + 2s2 - 8s3 = 24
for minimization problems
-2x2 +x3 + 2s2 - 4s3 = 8 (1)
! new obj. fn. value = old obj. fn. value – (new RHS – old RHS) × shadow price
x1+1.25x2 - .5s2 + 1.5s3 = 2
For the above example, as the allowable increases in RHS ranges are infinity for x2 +s4 = 5
each constraint, we can increase RHS of them as much as we want. But according to
allowable decreases, RHS of the first constraint can be decreased by 100 and that of
Analysis 1
second constraint by 1.
Suppose available finishing time changes from 20 * 20++, then we have the system:
Lets assume that new RHS value of the first constraint is 60.
As the change is within allowable range, we can use the first equation (max. z' = 60x1' + 30x2' + 20x3'
problem): 8x1' + 6x2' + x3' +s1' = 48

znew = 1001 + ( 60 - 100 ) 10 = 601.
4x1' + 2x2' + 1.5x3' +s2' = 20++
2x1' + 1.5x2' + .5x3' +s3' = 8
x2' +s4' = 5

38 39
or equivalently: Analysis 2
What happens if revenue from desks changes to $60+,? For small ,- revenue
z’ = 60x1’ + 30x2’ + 20x3’
increases by 2, (as we are making 2 desks currently). But how large an increase is
8x1’ + 6x2’ + x3’ +s1’ = 48
possible?
4x1’ + 2x2’ + 1.5x3’ +(s2’-+) = 20 The new revenue is:
2x1’ + 1.5x2’ + .5x3’ +s3’ = 8 z' = (60+,)x1+30x2+20x3 = z+,x1
x2’ +s4’ = 5 = (280-5x2-10s2-10s3)+,(2-1.25x2+.5s2-1.5s3)
That is z’,x1’,x2’,x3’,x4’,s1’,s2’-+,s3’,s4’ satisfy the original problem, and hence (1) = 280+2,-(5+1.25,)x2-(10-.5,)s2-(10+1.5,)s3
Substituting in:
So the top line in the final system would be:
z’ +5x2’ + 10(s2’-+) + 10s3’ = 280 z'+(5+1.25,)x2+(10-.5,)s2+(10+1.5,)s3 = 280+2,
-2x2’ +s1’ + 2(s2’-+) - 8s3’ = 24 Provided all terms in this row are $- we are still optimal. i.e. -4 #%,%# 20 the current
-2x2’ +x3’ 4s3’ production schedule is still optimal.

- 2(s2’-+) - = 8
x1’+1.25x2’ - .5(s2’-+) + 1.5s3’ = 2 Analysis 3

x2 ’ ’
+s4 = 5 If revenue from a non-basic variable changes, the revenue is
and thus z’ = 60x1+(30+,)x2+20x3 = z+,x2
z’ +5x2’ + 10s2’ + 10s3’ = 280+10+ = 280-5x2-10s2-10s3+,x2
-2x2’ +s1’ + 2s2’ - 8s3’ = 24+2+ = 280-(5-,)x2-10s2-10s3
-2x2’ +x3’ + 2s2’ - 4s3’ = 8 +2+ Thus the current solution is optimal for ,%# 5. But when ,%.%5 or the revenue per table
is increased past $35, it becomes better to produce tables. We say the reduced cost
x1’+1.25x2’ - .5s2’ + 1.5s3’ = 2-.5+
of tables is $5.00.
x2’ +s4’ = 5
For -4 #%+%# 4, the new system maximizes z’. In this range RHS are non-negative.
As + increases, revenue increases by 10+ and. we could pay up to $10 per hr. for
extra finishing labor and still increase revenue. Therefore, the shadow price of
finishing labor is $10 per hr. (This is valid for up to 4 extra hours or 4 fewer hours).

40 41
3.4 DUALITY 3.4.3 Finding the Dual of a Normal Min Problem
For an LP with m constraints and n variables if b is a one-dimensional vector of
3.4.1 The Dual Theorem length m, c and x are one-dimensional vectors of length n and A is a two-dimensional
Linear programming theory tells us that (provided the primal LP is feasible) the matrix with m rows and n columns the primal linear program (in matrix notation) is:
optimal value of the primal LP is equal to the optimal value of the dual LP minimize cx
subject to Ax >= b
3.4.2 Finding the Dual of a Normal Max Problem x >= 0
For an LP with m constraints and n variables if b is a one-dimensional vector of Associated with this primal linear program we have the dual linear program (involving
length m, c and x are one-dimensional vectors of length n and A is a two-dimensional a one-dimensional vector y of length m) given (in matrix notation) by:
matrix with m rows and n columns the primal linear program (in matrix notation) is: maximize by
maximize cx subject to yA <= c
subject to Ax <= b y >= 0
x >= 0 The tabular approach, mentioned above, can be used here similarly. If the primal is a
Associated with this primal linear program we have the dual linear program (involving normal min problem, we find it by reading down (Table 1); the dual is found by
a one-dimensional vector y of length m) given (in matrix notation) by: reading across in the table.
minimize by
subject to Ay >= c 3.4.4 Finding the Dual of a Nonnormal Max Problem
y >= 0 1. Fill in Table 1 so that the primal can be read across.
Table 1. Finding the dual of a problem 2. After making the following changes, the dual can be read down in the usual
max z fashion:
min w x1>0 x2>0 .... xn>0 a) If the ith primal constraint is a > constraint, the corresponding dual
. x1 x2 xn variable yi must satisfy yi < 0
y1>0 y1 a11 a12 .... a1n <b1 b) If the ith primal constraint is an equality constraint, the dual variable
y2>0 y2 a21 a22 .... a2n <b2 yi is now unrestricted in sign (urs).
.... .... .... .... .... .... c) If the ith primal variable is urs, the ith dual constraint will be an
ym>0 ym am1 am2 .... amn <bm
equality constraint
>c1 >c2 >cn
A tabular approach makes it easy to find the dual of an LP. If the primal is a normal 3.4.5 Finding the Dual of a Nonnormal Min Problem
max problem, it can be read across (Table 1); the dual is found by reading down the 1. Fill in Table 1 so that the primal can be read down.
table. 2. After making the following changes, the dual can be read across in the usual
fashion:
a) If the ith primal constraint is a < constraint, the corresponding dual
variable xi must satisfy xi < 0

42 43
b) If the ith primal constraint is an equality constraint, the dual variable
xi is now urs.
c) If the ith primal variable is urs, the ith dual constraint will be an
equality constraint 4. TRANSPORTATION PROBLEMS
3.4.6 An Example for Economic Interpretation (Dakota Furniture) In general, a transportation problem is specified by the following information:
Let x1, x2, x3 be the number of desks, tables and chairs produced. Let the weekly ! A set of m supply points form which a good is shipped. Supply point i can
profit be $z. Then, we must supply at most si units.
max z = 60x1+30x2+20x3 ! A set of n demand points to which the good is shipped. Demand point j must
s.t. 8x1+ 6x2+ x3 < 48 (Lumber constraint) receive at least dj units of the shipped good.
4x1+ 2x2+1.5x3 < 20 (Finishing hour constraint) ! Each unit produced at supply point i and shipped to demand point j incurs a
2x1+1.5x2+0.5x3 < 8 (Carpentry hour constraint) variable cost of cij.
x1,x2,x3 > 0 The relevant data can be formulated in a transportation tableau:
Demand Demand Demand
..... SUPPLY
point 1 point 2 point n
Dual Supply c11 c12 c1n
s1
Suppose an entrepreneur wants to purchase all of Dakota’s resources. point 1
Supply c21 c22 c2n
In the dual problem y1, y2, y3 are the resource prices (price paid for one board ft of s2
point 2
lumber, one finishing hour, and one carpentry hour).
.....
$w is the cost of purchasing the resources.
Supply cm1 cm2 cmn
Resource prices must be set high enough to induce Dakota to sell. i.e. total sm
point m
purchasing cost equals total profit. DEMAND d1 d2 dn
min w = 48y1+ 20y2+ 8y3
s.t. 8y1+ 4y2+ 2y3 > 60 (Desk constraint) If total supply equals total demand then the problem is said to be a balanced
6y1+ 2y2+1.5y3 > 30 (Table constraint) transportation problem.
y1+1.5y2+0.5y3 > 20 (Chair constraint)

y1,y2,y3 > 0

44 45
4.1 FORMULATING TRANSPORTATION PROBLEMS Answer
Formulation of the transportation problem
4.1.1 LP Representation City 1 City 2 City 3 City 4 SUPPLY
8 6 10 9
Let xij = number of units shipped from supply point i to demand point j Plant 1 35
then the general LP representation of a transportation problem is 9 12 13 7
Plant 2 50
min /i%/j cij xij 14 9 16 5
Plant 3 40
s.t. /j xij < si (i=1,2, ..., m) Supply constraints
DEMAND 45 20 30 30 125
/i xij > dj (j=1,2, ..., n) Demand constraints Total supply & total demand both equal 125: “balanced transport’n problem”.
xij > 0
If a problem has the constraints given above and is a maximization problem, it is still Representation of the problem as a LP model
a transportation problem. xij: number of (million) kwh produced at plant i and sent to city j.
min z = 8 x11 + 6 x12 + 10 x13 + 9 x14 + 9 x21 + 12 x22 + 13 x23 + 7 x24 + 14
4.1.2 Powerco Example for Balanced Transportation Problem x31 + 9 x32 + 16 x33 + 5 x34
Powerco has three electric power plants that supply the needs of four cities. Each s.t. x11 + x12 + x13 + x14 < 35 (supply constraints)
power plant can supply the following numbers of kwh of electricity: plant 1, 35 million; x21 + x22 + x23 + x24 < 50
plant 2, 50 million; and plant 3, 40 million. The peak power demands in these cities x31 + x32 + x33 + x34 < 40
as follows (in kwh): city 1, 45 million; city 2, 20 million; city 3, 30 million; city 4, 30 x11 + x21 + x31 > 45 (demand constraints)
million. The costs of sending 1 million kwh of electricity from plant to city is given in x12 + x22 + x32 > 20
the table below. To minimize the cost of meeting each city’s peak power demand, x13 + x23 + x33 > 30
formulate a balanced transportation problem in a transportation tableau and x14 + x24 + x34 > 30
represent the problem as a LP model. xij > 0 (i = 1, 2, 3; j = 1, 2, 3, 4)
To
From City 1 City 2 City 3 City 4 4.1.3 Balancing an Unbalanced Transportation Problem
Plant 1 $8 $6 $10 $9
Plant 2 $9 $12 $13 $7
Plant 3 $14 $9 $16 $5 Excess Supply
If total supply exceeds total demand, we can balance a transportation problem by
creating a dummy demand point that has a demand equal to the amount of excess
supply.
Since shipments to the dummy demand point are not real shipments, they are
assigned a cost of zero.
These shipments indicate unused supply capacity.

46 47
Unmet Demand 4.1.5 Modified Powerco Example for Unmet Demand
If total supply is less than total demand, actually the problem has no feasible solution. Suppose that demand for city 1 is 50 million kwh. For each million kwh of unmet
To solve the problem it is sometimes desirable to allow the possibility of leaving demand, there is a penalty of 80$. Formulate a balanced transportation problem.
some demand unmet. In such a situation, a penalty is often associated with unmet
demand. This means that a dummy supply point should be introduced. Answer
We would add a dummy supply point having a supply of 5 million kwh representing
4.1.4 Modified Powerco Example for Excess Supply shortage.
Suppose that demand for city 1 is 40 million kwh. Formulate a balanced City 1 City 2 City 3 City 4 SUPPLY
8 6 10 9
transportation problem. Plant 1 35
9 12 13 7
Plant 2 50
Answer
14 9 16 5
Total demand is 120, total supply is 125. Plant 3 40
To balance the problem, we would add a dummy demand point with a demand of 125 Dummy 80 80 80 80
5
(Shortage)
– 120 = 5 million kwh.
DEMAND 50 20 30 30 130
From each plant, the cost of shipping 1 million kwh to the dummy is 0.
For details see Table 4.
Table 4. Transportation Tableau for Excess Supply
City 1 City 2 City 3 City 4 Dummy SUPPLY
8 6 10 9 0
Plant 1 35
9 12 13 7 0
Plant 2 50
14 9 16 5 0
Plant 3 40
DEMAND 40 20 30 30 5 125

48 49
4.2 FINDING BFS FOR TRANSPORT’N PROBLEMS 4.2.1 Northwest Corner Method
For a balanced transportation problem, the general LP representation may be written We begin in the upper left corner of the transportation tableau and set x11 as large as
as: possible (clearly, x11 can be no larger than the smaller of s1 and d1).
min /i%/j cij xij ! If x11=s1, cross out the first row of the tableau. Also change d1 to d1-s1.
! If x11=d1, cross out the first column of the tableau. Change s1 to s1-d1.
s.t. /j xij = si (i=1,2, ..., m) Supply constraints
! If x11=s1=d1, cross out either row 1 or column 1 (but not both!).
/i xij = dj (j=1,2, ..., n) Demand constraints o If you cross out row, change d1 to 0.
xij > 0 o If you cross out column, change s1 to 0.
To find a bfs to a balanced transportation problem, we need to make the following Continue applying this procedure to the most northwest cell in the tableau that does
important observation: not lie in a crossed out row or column.
If a set of values for the xij’s satisfies all but one of the constraints of a balanced Eventually, you will come to a point where there is only one cell that can be assigned
transportation problem, the values for the xij’s will automatically satisfy the other a value. Assign this cell a value equal to its row or column demand, and cross out
constraint. both the cell’s row or column.
This observation shows that when we solve a balanced transportation, we may omit A bfs has now been obtained.
from consideration any one of the problem’s constraints and solve an LP having
m+n-1 constraints. We arbitrarily assume that the first supply constraint is omitted Example:
from consideration. For example consider a balanced transportation problem given below (We omit the
In trying to find a bfs to the remaining m+n-1 constraints, you might think that any costs because they are not needed to find a bfs!).
collection of m+n-1 variables would yield a basic solution. But this is not the case: 5
If the m+n-1 variables yield a basic solution, the cells corresponding to a set of m+n-
1
1 variables contain no loop.
3
An ordered sequence of at least four different cells is called a loop if
! Any two consecutives cells lie in either the same row or same column 2 4 2 1
! No three consecutive cells lie in the same row or column Total demand equals total supply (9): this is a balanced transport’n problem.
! The last cell in the sequence has a row or column in common with the first cell
2 3
in the sequence
1
There are three methods that can be used to find a bfs for a balanced transportation
problem: 3
1. Northwest Corner method X 4 2 1
2. Minimum cost method
3. Vogel’s method

50 51
4.2.2 Minimum Cost Method
2 3 X
Northwest Corner method does not utilize shipping costs, so it can yield an initial bfs
1
that has a very high shipping cost. Then determining an optimal solution may require
3 several pivots.
X 1 2 1 To begin the minimum cost method, find the variable with the smallest shipping cost
(call it xij). Then assign xij its largest possible value, min {si, dj}.
2 3 X As in the NWC method, cross out row i or column j and reduce the supply or demand
of the noncrossed-out of row or column by the value of xij.
1 X
Continue like NWC method (instead of assigning upper left corner, the cell with the
3
minimum cost is assigned). See Northwest Corner Method for the details!
X 0 2 1
2 3 5 6
5
2 3 X
2 1 3 5
10
1 X
3 8 4 6
0 2 1 3 15
X 0 2 1 12 8 4 6
NWC method assigned values to m+n-1 (3+4-1 = 6) variables. The variables chosen 2 3 5 6
5
by NWC method can not form a loop, so a bfs is obtained.
2 1 3 5
2
8
3 8 4 6
15
12 X 4 6
2 3 5 6
5
2 1 3 5
X
2 8
3 8 4 6
15
10 X 4 6

52 53
4.2.3 Vogel’s Method
2 3 5 6 Begin by computing for each row and column a penalty equal to the difference
X
5
between the two smallest costs in the row and column.
2 1 3 5
X
2 8 Next find the row or column with the largest penalty.
3 8 4 6 Choose as the first basic variable the variable in this row or column that has the
15
smallest cost.
5 X 4 6
As described in the NWC method, make this variable as large as possible, cross out
2 3 5 6 row or column, and change the supply or demand associated with the basic variable
X
5 (See Northwest Corner Method for the details!).
2 1 3 5
X Now recomputed new penalties (using only cells that do not lie in a crossed out row
2 8
3 8 4 6 or column), and repeat the procedure until only one uncrossed cell remains. Set this
10
5 4 6
variable equal to the supply or demand associated with the variable, and cross out
5 X 4 6
the variable’s row and column.
Row
Supply
penalty
6 7 8
10 7-6=1
15 80 78
15 78-15=63
Demand 15 5 5
Column
15-6=9 80-7=73 78-8=70
penalty
Row
Supply
penalty
6 7 8
5 8-6=2
5
15 80 78
15 78-15=63
Demand 15 X 5
Column
15-6=9 - 78-8=70
penalty

54 55
Row 4.3 THE TRANSPORTATION SIMPLEX METHOD
Supply
penalty
6 7 8
X -
5 5 4.3.1 Steps of the Method
15 80 78 1. If the problem is unbalanced, balance it
15 -
2. Use one of the methods to find a bfs for the problem
Demand 15 X 0
3. Use the fact that u1 = 0 and ui + vj = cij for all basic variables to find the u’s and v’s
Column
15-6=9 - - for the current bfs.
penalty
4. If ui + vj – cij < 0 for all nonbasic variables, then the current bfs is optimal. If this is
6 7 8 not the case, we enter the variable with the most positive ui + vj – cij into the basis
X
5 5 using the pivoting procedure. This yields a new bfs.
15 80 78
15 5. Using the new bfs, return to Steps 3 and 4.
15 0
Demand 15 X 0
For a maximization problem, proceed as stated, but replace Step 4 by the following
step:
If ui + vj – cij > 0 for all nonbasic variables, then the current bfs is optimal.
Otherwise, enter the variable with the most negative ui + vj – cij into the basis using
the pivoting procedure. This yields a new bfs.
Pivoting procedure
1. Find the loop (there is only one possible loop!) involving the entering variable
(determined at step 4 of the transport’n simplex method) and some of the basic
variables.
2. Counting only cells in the loop, label those that are an even number (0, 2, 4, and
so on) of cells away from the entering variable as even cells. Also label those that
are an odd number of cells away from the entering variable as odd cells.
3. Find the odd cell whose variable assumes the smallest value. Call this value 0.
The variable corresponding to this odd cell will leave the basis. To perform the
pivot, decrease the value of each odd cell by 0 and increase the value of each
even cell by 0. The values of variables not in the loop remain unchanged. The
pivot is now complete. If 0 = 0, the entering variable will equal 0, and odd variable
that has a current value of 0 will leave the basis.

56 57
4.3.2 Solving Powerco Example by Transportation Simplex x33 would leave the basis. New bfs is shown at the following table:
The problem is balanced (total supply equals total demand).
ui/vj 8 11 12 7 SUPPLY
When the NWC method is applied to the Powerco example, the bfs in the following
8 6 10 9
table is obtained (check: there exist m+n–1=6 basic variables). 0 35
35
City 1 City 2 City 3 City 4 SUPPLY 9 12 13 7
1 50
8 6 10 9 10 10 30
Plant 1 35 14 9 16 5
35 -2 40
9 12 13 7 10% % 30
Plant 2 50
10 20 20 DEMAND 45 20 30 30 125
14 9 16 5
Plant 3 40
10 30 !12 = 5, !13 = 2, !14 = -2, !24 = 1, !31 = -8, !33 = -6
DEMAND 45 20 30 30 125
u1 = 0 Since !12 is the most positive one, we would next enter x12 into the basis.
u1 + v1 = 8 yields v1 = 8 The loop involving x12 is (1,2)-(2,2)-(2,1)-(1,1). 0 = 10 (see table)
u2 + v1 = 9 yields u2 = 1 City 1 City 2 City 3 City 4 SUPPLY

8 6 10 9
u2 + v2 = 12 yields v2 = 11 Plant 1 35
35–0 0
u2 + v3 = 13 yields v3 = 12 9 12 13 7
Plant 2 50
10+0 10–0 30
u3 + v3 = 16 yields u3 = 4
14 9 16 5
Plant 3 40
u3 + v4 = 5 yields v4 = 1 10% % 30
For each nonbasic variable, we now compute !ij = ui + vj – cij DEMAND 45 20 30 30 125
!12 = 0 + 11 – 6 = 5
x22 would leave the basis. New bfs is shown at the following table:
!13 = 0 + 12 – 10 = 2
!14 = 0 + 1 – 9 = -8 ui/vj 8 6 12 2 SUPPLY
!24 = 1 + 1 – 7 = -5 8 6 10 9
0 35
25 10
!31 = 4 + 8 – 14 = -2 9 12 13 7
1 50
!32 = 4 + 11 – 9 = 6 20 30
14 9 16 5
Since !32 is the most positive one, we would next enter x32 into the basis: Each unit of 3 40
10% % 30
x32 that is entered into the basis will decrease Powerco’s cost by $6. DEMAND 45 20 30 30 125
The loop involving x32 is (3,2)-(3,3)-(2,3)-(2,2). 0 = 10 (see table)
!13 = 2, !14 = -7, !22 = -5, !24 = -4, !31 = -3, !33 = -1
City 1 City 2 City 3 City 4 SUPPLY
8 6 10 9
Plant 1 35
35 Since !13 is the most positive one, we would next enter x13 into the basis.
9 12 13 7
Plant 2 50
10 20–0 20+0
14 9 16 5
Plant 3 40
0% 10–0% 30
DEMAND 45 20 30 30 125

58 59
The loop involving x13 is (1,3)-(2,3)-(2,1)-(1,1). 0 = 25 (see table) 4.4 TRANSSHIPMENT PROBLEMS
City 1 City 2 City 3 City 4 SUPPLY Sometimes a point in the shipment process can both receive goods from other points
8 6 10 9 and send goods to other points. This point is called as transshipment point through
Plant 1 35
25–0 10 0
9 12 13 7 which goods can be transshipped on their journey from a supply point to demand
Plant 2 50
20+0 30–0 point.
14 9 16 5
Plant 3 40 Shipping problem with this characteristic is a transshipment problem.
10% % 30
The optimal solution to a transshipment problem can be found by converting this
DEMAND 45 20 30 30 125
transshipment problem to a transportation problem and then solving this
x11 would leave the basis. New bfs is shown at the following table: transportation problem.
ui/vj 6 6 10 2 SUPPLY
Remark
8 6 10 9
0 35 As stated in “Formulating Transportation Problems”, we define a supply point to be
10 25
9 12 13 7 a point that can send goods to another point but cannot receive goods from any other
3 50
45 5
14 9 16 5 point.
3 40
10% % 30 Similarly, a demand point is a point that can receive goods from other points but
DEMAND 45 20 30 30 125 cannot send goods to any other point.
!11 = -2, !14 = -7, !22 = -3, !24 = -2, !31 = -5, !33 = -3
4.4.1 Steps
Since all !ij’s are negative, an optimal solution has been obtained. 1. If the problem is unbalanced, balance it
Let s = total available supply (or demand) for balanced problem
Report 2. Construct a transportation tableau as follows
45 million kwh of electricity would be sent from plant 2 to city 1. A row in the tableau will be needed for each supply point and transshipment point
10 million kwh of electricity would be sent from plant 1 to city 2. Similarly, 10 million A column will be needed for each demand point and transshipment point
kwh of electricity would be sent from plant 3 to city 2. Each supply point will have a supply equal to its original supply
25 million kwh of electricity would be sent from plant 1 to city 3. 5 million kwh of Each demand point will have a demand equal to its original demand
electricity would be sent from plant 2 to city 3. Each transshipment point will have a supply equal to “that point’s original supply +
30 million kwh of electricity would be sent from plant 3 to city 4 and s”
Total shipping cost is: Each transshipment point will have a demand equal to “that point’s original
z = .9 (45) + 6 (10) + 9 (10) + 10 (25) + 13 (5) + 5 (30) = $ 1020 demand + s”
3. Solve the transportation problem

60 61
4.4.2 Kuruoglu Example Table 2. Representation of Transship’t Problem as Balanced Transp’n Problem
(Based on Winston 7.6.) Ankara Eskisehir Istanbul Izmir Dummy Supply
Kuruoglu manufactures refrigerators at two factories, one in Malatya and one in G.Antep. 8 13 25 28 0
Malatya 150
The Malatya factory can produce up to 150 refrigerators per day, and the G.Antep factory
15 12 26 25 0
G.Antep 200
can produce up to 200 refrigerators per day. Refrigerators are shipped by air to customers in
Istanbul and Izmir. The customers in each city require 130 refrigerators per day. Because of 0 6 16 17 0
Ankara 350
the deregulation of air fares, Kuruoglu believes that it may be cheaper to first fly some 6 0 14 16 0
Eskisehir 350
refrigerators to Ankara or Eskisehir and then fly them to their final destinations. The costs of
flying a refrigerator are shown at the below table. Kuruoglu wants to minimize the total cost of Demand 350 350 130 130 90
shipping the required refrigerators to its customers.

Step 3. Solving the transportation problem
Table 1. Shipping Costs for Transshipment
Table 3. Optimal solution for Transp’n Problem
MTL To
From Malatya G.Antep Ankara Eskisehir Istanbul Izmir Ankara Eskisehir Istanbul Izmir Dummy Supply
Malatya 0 - 8 13 25 28 8 13 25 28 0
G.Antep - 0 15 12 26 25 Malatya 150
130 20
Ankara - - 0 6 16 17
15 12 26 25 0
Eskisehir - - 6 0 14 16 G.Antep 200
130 70
Istanbul - - - - 0 -
0 6 16 17 0
Izmir - - - - - 0 Ankara 350
220 130
6 0 14 16 0
Eskisehir 350
Answer: 350
In this problem Ankara and Eskisehir are transshipment points. Demand 350 350 130 130 90 1050
Step 1. Balancing the problem

Report:
Total supply = 150 + 200 = 350
Kuruoglu should produce 130 refrigerators at Malatya, ship them to Ankara, and
Total demand = 130 + 130 = 260
transship them from Ankara to Istanbul.
Dummy’s demand = 350 – 260 = 90
The 130 refrigerators produced at G.Antep should be shipped directly to "zmir.
s = 350 (total available supply or demand for balanced problem)
The total shipment is 6370 MTL.
Step 2. Constructing a transportation tableau

Remark: In interpreting the solution to this problem, we simply ignore the shipments
Transshipment point’s demand = Its original demand + s = 0 + 350 = 350
to the dummy and from a point itself.
Transshipment point’s supply = Its original supply + s = 0 + 350 = 350

62 63
4.5 ASSIGNMENT PROBLEMS Remarks
There is a special case of transportation problems where each supply point should 1. To solve an assignment problem in which the goal is to maximize the objective
be assigned to a demand point and each demand should be met. This certain class function, multiply the profits matrix through by –1 and solve the problem as a
of problems is called as “assignment problems”. For example determining which minimization problem.
employee or machine should be assigned to which job is an assignment problem. 2. If the number of rows and columns in the cost matrix are unequal, the assignment
problem is unbalanced. Any assignment problem should be balanced by the
4.5.1 LP Representation addition of one or more dummy points before it is solved by the Hungarian
An assignment problem is characterized by knowledge of the cost of assigning each method.
supply point to each demand point: cij
On the other hand, a 0-1 integer variable xij is defined as follows Steps
xij = 1 if supply point i is assigned to meet the demands of demand point j 1. Find the minimum cost each row of the m*m cost matrix.
xij = 0 if supply point i is not assigned to meet the demands of point j 2. Construct a new matrix by subtracting from each cost the minimum cost in its row
In this case, the general LP representation of an assignment problem is 3. For this new matrix, find the minimum cost in each column
min /i%/j cij xij 4. Construct a new matrix (reduced cost matrix) by subtracting from each cost the
minimum cost in its column
s.t. /j xij = 1 (i=1,2, ..., m) Supply constraints
5. Draw the minimum number of lines (horizontal and/or vertical) that are needed to
/i xij = 1 (j=1,2, ..., n) Demand constraints cover all the zeros in the reduced cost matrix. If m lines are required, an optimal
xij = 0 or xij = 1 solution is available among the covered zeros in the matrix. If fewer than m lines
are needed, proceed to next step
4.5.2 Hungarian Method 6. Find the smallest nonzero cost (k) in the reduced cost matrix that is uncovered by
Since all the supplies and demands for any assignment problem are integers, all the lines drawn in Step 5
variables in optimal solution of the problem must be integers. Since the RHS of each 7. Subtract k from each uncovered element of the reduced cost matrix and add k to
constraint is equal to 1, each xij must be a nonnegative integer that is no larger than each element that is covered by two lines. Return to Step 5
1, so each xij must equal 0 or 1.

Ignoring the xij = 0 or xij = 1 restrictions at the LP representation of the assignment
problem, we see that we confront with a balanced transportation problem in which
each supply point has a supply of 1 and each demand point has a demand of 1.
However, the high degree of degeneracy in an assignment problem may cause the
Transportation Simplex to be an inefficient way of solving assignment problems.
For this reason and the fact that the algorithm is even simpler than the Transportation
Simplex, the Hungarian method is usually used to solve assignment problems.

64 65
4.5.3 Flight Crew Example Table 4. Reduced cost matrix with lines covering zeros
(Based on Winston 7.5.) 0 2 4 6
0 10 4 1
Four captain pilots (Selçuk, Serkan, Ümit, Volkan) has evaluated four flight officers 4 5 0 4
9 0 3 0
(Tuncay, Önder, Servet, Kemal) according to perfection, adaptation, morale
Step 6 & 7. The smallest uncovered cost equals 1. We now subtract 1 from each
motivation in a 1-20 scale (1: very good, 20: very bad). Evaluation grades are given
uncovered cost, add 1 to each twice-covered cost, and obtain Table 6.
in Table 1. Flight Company wants to assign each flight officer to a captain pilot
Table 6. Resulting matrix
according to these evaluations. Determine possible flight crews.
0 1 3 5
Table 1. Times (necessary weeks) for Papers 0 9 3 0
5 5 0 4
Tuncay Önder Servet Kemal 10 0 3 0
Selçuk 2 4 6 10
Four lines are now required to cover all the zeros: An optimal s9olution is available.
Serkan 2 12 6 5
Ümit 7 8 3 9 Observe that the only covered 0 in column 3 is x33, and in column 2 is x42. As row 5
Volkan 14 5 8 7
can not be used again, for column 4 the remaining zero is x24. Finally we choose x11.
Answer:
Report:
Step 1. For each row in Table 1, we find the minimum cost: 2, 2, 3, and 5
CP Selçuk should fly with FO Tuncay; CP Serkan should fly with FO Kemal; CP Ümit
respectively
should fly with FO Servet; and CP Volkan should fly with FO Önder.
Step 2 & 3. We subtract the row minimum from each cost in the row obtaining Table
2. For this new matrix, we find the minimum cost in each column
Table 2. Cost matrix after row minimums are subtracted
0 2 4 8
0 10 4 3
4 5 0 6
9 0 3 2
Column minimum 0 0 0 2
Step 4. We now subtract the column minimum from each cost in the column
obtaining reduced cost matrix in Table 3.
Table 3. Reduced cost matrix
0 2 4 6
0 10 4 1
4 5 0 4
9 0 3 0
Step 5. As shown in Table 4, lines through row 3, row 4, and column 1 cover all the
zeros in the reduced cost matrix. The minimum number of lines for this operation is 3.
Since fewer than four lines are required to cover all the zeros, solution is not optimal:
we proceed to next step.

66 67
5.1 FORMULATING IP
5.1.1 Budgeting Problems
5.1.1.1 Capital Budgeting Example
5. INTEGER PROGRAMMING Suppose we wish to invest $14,000. We have identified four investment
opportunities. Investment 1 requires an investment of $5,000 and has a present
When formulating LP’s we often found that, strictly, certain variables should have value (a time-discounted value) of $8,000; investment 2 requires $7,000 and has a
been regarded as taking integer values but, for the sake of convenience, we let them value of $11,000; investment 3 requires $4,000 and has a value of $6,000; and
take fractional values reasoning that the variables were likely to be so large that any investment 4 requires $3,000 and has a value of $4,000. Into which investments
fractional part could be neglected. While this is acceptable in some situations, in should we place our money so as to maximize our total present value?
many cases it is not, and in such cases we must find a numeric solution in which the
variables take integer values. Answer
Problems in which this is the case are called integer programs (IP's) and the subject As in LP, our first step is to decide on our variables. This can be much more difficult
of solving such programs is called integer programming (also referred to by the in IP because there are very clever or tricky ways to use integrality restrictions.
initials IP).
IP's occur frequently because many decisions are essentially discrete (such as In this case, we will use a 0-1 variable xj for each investment.
yes/no, do/do not) in that one or more options must be chosen from a finite set of If xj is 1 then we will make investment j.
alternatives. If it is 0, we will not make the investment.
An integer programming problem in which all variables are required to be integer is This leads to the 0-1 programming problem:
called a pure integer programming problem. Maximize 8 x1 + 11 x2 + 6 x3 + 4 x4
If some variables are restricted to be integer and some are not then the problem is a Subject to 5 x1 + 7 x2 + 4 x3 + 3 x4 < 14
mixed integer programming problem. xj = 0 or 1 j = 1, … 4
The case where the integer variables are restricted to be 0 or 1 comes up surprising
often. Such problems are called pure (mixed) 0-1 programming problems or pure Now, a straightforward “bang for buck” (taking ratios of objective coefficient over
(mixed) binary integer programming problems. constraint coefficient) suggests that investment 1 is the best choice.
Ignoring integrality constraints, the optimal linear programming solution is
x1 = 1, x2 = 1, x3 = 0.5, and x4 = 0 for a value of $22,000.
Unfortunately, this solution is not integral. Rounding x3 down to 0 gives a feasible
solution with a value of $19,000.
There is a better integer solution, however, of x1 = 0, x2 = x3 = x4 = 1 for a value of
$21,000.
This example shows that rounding does not necessarily give an optimal value.

68 69
5.1.1.2 Multiperiod Example 5.1.1.3 Capital Budgeting Extension
There are four possible projects, which each run for 3 years and have the following There are a number of additional constraints we might want to add. Logical
characteristics. restrictions can be enforced using 0-1 variables.
Capital requirements For instance, consider the following constraints:
Project Return Year1 Year2 Year3
1 0.2 0.5 0.3 0.2 We can only make two investments
2 0.3 1 0.5 0.2 x1 + x2 + x3 + x4 < 2
3 0.5 1.5 1.5 0.3
Any choice of 3 or 4 investments will have x1 + x2 + x3 + x4 > 3
4 0.1 0.1 0.4 0.1
Available capital 3.1 2.5 0.4
If investment 2 is made, investment 4 must also be made
Which projects would you choose in order to maximize the total return? x2 < x4 or x2 – x4 < 0
If x2 is 1, then x4 is also 1 as we desire; if x2 is 0, then there is no restriction for x4 (x4
Answer is 0 or 1)
xj is 1 if we decide to do project j; xj is 0 otherwise (i.e. not do project j).
This leads to the 0-1 programming problem: If investment 1 is made, investment 3 cannot be made
Maximize 0.2 x1 + 0.3 x2 + 0.5 x3 + 0.1 x4 x1 + x3 < 1
Subject to 0.5 x1 + 1 x2 + 1.5 x3 + 0.1 x4 < 3.1 If x1 is 1, then x3 is 0 as we desire; if x1 is 0, then there is no restriction for x3 (x3 is 0
0.3 x1 + 0.8 x2 + 1.5 x3 + 0.4 x4 < 2.5 or 1)
0.2 x1 + 0.2 x2 + 0.3 x3 + 0.1 x4 < 0.4
xj = 0 or 1 j = 1, … 4 Either investment 1 or investment 2 must be done
x1 + x2 = 1
Remarks If x1 is 1, then x2 is 0 (only investment 1 is done); if x1 is 0, then x2 is 1 (only
You will note that the objective and constraints are linear. In this course we deal only investment 2 is done)
with linear integer programs (IP's with a linear objective and linear constraints). It is
plain though that there do exist non-linear integer programs - these are, however,
outside the scope of this course.
Whereas before in formulating LP's if we had integer variables we assumed that we
could ignore any fractional parts it is clear that we cannot do so in this problem e.g.
what would be the physical meaning of a numeric solution with x1=0.4975?

70 71
5.1.2 Knapsack Problems 5.1.3 Fixed Charge Problems
Any IP that has only one constraint is referred to as a knapsack problem. An important trick can be used to formulate many production and location problems
Furthermore, the coefficients of this constraint and the objective are all non-negative. as IP.
For example, the following is a knapsack problem: Here Gandhi example is given as a production problem and Lockbox example is
Maximize 8 x1 + 11 x2 + 6 x3 + 4 x4 given as a location problem.
Subject to 5 x1 + 7 x2 + 4 x3 + 3 x4 < 14
xj = 0 or 1 j = 1, … 4 5.1.3.1 Gandhi Example (Production)
(Winston 9.2., p. 470)
The traditional story is that there is a knapsack (here of capacity 14). There are a Gandhi Co makes shirts, shorts, and pants using the limited labor and cloth
number of items (here there are four items), each with a size and a value (here item described below. In addition, the machinery to make each product must be rented.
2 has size 7 and value 11). Shirts Shorts Pants Total Avail.
The objective is to maximize the total value of the items in the knapsack. Labor (hrs/wk) 3 2 6 150
Cloth (m2/wk) 4 3 4 160
Knapsack problems are nice because they are (usually) easy to solve.
Rent for machine ($/wk) 200 150 100
Variable unit cost 6 4 8
Sale Price 12 8 15
Answer
Let xj be number of clothing produced.
Let yj be 1 if any clothing j is manufactured and 0 otherwise.
Profit = Sales revenue – Variable Cost – Costs of renting machinery

For example the profit from shirts is z1 = ( 12 – 6 ) x1 – 200 y1.
Since supply of labor and cloth is limited, Gandhi faces two constraints.
To ensure xj > 0 forces yj = 1, we include the additional constraints

xj < Mj yj
From the cloth constraint at most 40 shirts can be produced (M1 = 40), so the
additional constraint for shirts is not an additional limit on x1 (If M1 were not chosen
large (say M1 = 10), then the additional constraint for shirts would unnecessarily
restrict the value of x1).
From the cloth constraint at most 53 shorts can be produced (M2 = 53) and from the
labor constraint at most 25 pants can be produced (M3 = 25).

72 73
We thus get the mixed (binary) integer problem: 5.1.3.2 The Lockbox Example (Location)
(Adapted to Winston 9.2., p. 473)
max 6 x1 + 4 x2 + 7 x3 – 200 y1 – 150 y2 – 100 y3 Consider a national firm that receives checks from all over the United States. There
st 3 x1 + 2 x2 + 6 x3 < 150 (Labor constraint) is a variable delay from when the check is postmarked (and hence the customer has
4 x1 + 3 x2 + 4 x3 < 160 (Cloth constraint) met her obligation) and when the check clears (the firm can use the money). It is in
x1 < 40 y1 (Shirt production and machinery constraint) the firm's interest to have the check clear as quickly as possible since then the firm
x2 < 53 y2 (Short production and machinery constraint) can use the money. To speed up this clearing, firms open offices (lockboxes) in
x3 < 25 y3 (Pant production and machinery constraint) different cities to handle the checks.
x1, x2, x3 > 0 and integer For example, suppose we receive payments from four regions (West, Midwest, East,
y1, y2, y3 = 0 or 1 and South). The average daily value from each region is as follows: $70,000 from the
West, $50,000 from the Midwest, $60,000 from the East, and $40,000 from the
The optimal solution to the Gandhi problem is z = $ 75, x3 = 25, y3 = 1. South. We are considering opening lockboxes in L.A., Chicago, New York, and/or
Thus, they should produce 25 pants and rent a pants machinery. Atlanta. Operating a lockbox costs $50,000 per year. The average days from mailing
to clearing is given in the table.
LA Chicago NY Atlanta
West 2 6 8 8
Midwest 6 2 5 5
East 8 5 2 5
South 8 5 5 2
Which lockboxes should we open? Formulate an IP that we can use to minimize the
sum of costs due to lost interest and lockbox operations. Assume that each region
must send all its money to a single city and investment rate is 20%.
Answer
First we must calculate the losses due to lost interest for each possible assignment.
For example, if the West sends to New York, then on average there will be $560,000
(= 8 * $70.000) in process on any given day. Assuming an investment rate of 20%,
this corresponds to a yearly loss of $112,000. We can calculate the losses for the
other possibilities in a similar fashion to get the following table.
LA Chicago NY Atlanta
West 28 84 112 112
Midwest 60 20 50 50
East 96 60 24 60
South 64 40 40 16
Let yj be a 0-1 variable that is 1 if lockbox j is opened and 0 if it is not.

Let xij be 1 if region i sends to lockbox j; and 0 otherwise.

74 75
Our objective is to minimize our total yearly costs. This is: For instance, instead of four constraints of the form
28 x11 + 84 x12 + 112 x13 + … + 50 y1 + 50 y2 + 50 y3 + 50 y4 x1j + x2j + x3j + x4j < M yj
consider the sixteen constraints of the form:
One set of constraint is that each region must be assigned to one lockbox: xij < yj i = 1, 2, 3, 4; j = 1, 2, 3, 4
SUM [j=1 to n] xij = 1 for all i
(SUM[j=1 to n] should be read as "sum over all integer values of j from 1 to n These constraints also force a region to only use open lockboxes.
inclusive") (Suppose yj is 0, so by using four corresponding constraints all of x1j, x2j, x3j, and x4j
must also be 0. If yj is 1, then there is no restriction on the x values.
A more difficult set of constraint is that a region can only be assigned to an open Another point of view: If xij = 1, then yj = 1 as desired. Also if x1j = x2j = x3j = x4j = 0,
lockbox. This can be written as then there is no restriction on the y values. The act of minimizing costs will result yj =
x1j + x2j + x3j + x4j < M yj 0.
M is any number that should be at least 4 as there are four regions.
(Suppose we do not open LA lockbox; then y1 is 0, so all of x11, x21, x31, and x41 must It might seem that a larger formulation (twenty variables and twenty constraints) is
also be 0. If y1 is 1, then there is no restriction on the x values.) less efficient and therefore should be avoided. This is not the case! If we solve the
linear program with the above constraints, we have an integer solution as an optimal
For this problem we would have twenty variables (four y variables, sixteen x LP solution, which must therefore be optimal!
variables) and eight constraints. This gives the following 0-1 linear program:
Min 28 x11 + 84 x12 + 112 x13 + 112 x14
+ 60 x21 + 20 x22 + 50 x23 + 50 x24
+ 96 x31 + 60 x32 + 24 x33 + 60 x34
+ 64 x41 + 40 x42 + 40 x43 + 16 x44
+ 50 y1 + 50 y2 + 50 y3 + 50 y4
st x11 + x12 + x13 + x14 = 1
x21 + x22 + x23 + x24 = 1
x31 + x32 + x33 + x34 = 1
x41 + x42 + x43 + x44 = 1
x11 + x21 + x31 + x41 < 4y1
x12 + x22 + x32 + x42 < 4y2
x13 + x23 + x33 + x43 < 4y3
x14 + x24 + x34 + x44 < 4y4
All xij and yj = 0 or 1
There are other formulations, however.

76 77
5.1.4 Set Covering Problems Answer
To illustrate this model, consider the following location problem: A county is reviewing We can create one variable xj for each city j.
the location of its fire stations. The county is made up of a number of cities, as This variable will be 1 if we place a station in the city, and will be 0 otherwise. This
illustrated in the following figure. leads to the following formulation
Min x1 + x2 + x3 + x4 + x5 + x6 + x7 + x8 + x9 + x10 + x11
st x1 + x2 + x3 + x4 > 1 (city 1)
x1 + x2 + x3 + x5 > 1 (city 2)
x1 + x2 + x3 + x4 + x5 + x6 > 1 (city 3)
x1 + x3 + x4 + x6 + x7 > 1 (city 4)
x2 + x3 + x5 + x6 + x8 + x9 > 1 (city 5)
x3 + x4 + x5 + x6 + x7 + x8 > 1 (city 6)
x4 + x6 + x7 + x8 > 1 (city 7)
x5 + x6 + x7 + x8 + x9 + x10 > 1 (city 8)
x5 + x8 + x9 + x10 + x11 > 1 (city 9)
+ x8 + x9 + x10 + x11 > 1 (city 10)
+ x9 + x10 + x11 > 1 (city 11)
All xj = 0 or 1
A fire station can be placed in any city. It is able to handle the fires for both its city
and any adjacent city (any city with a non-zero border with its home city). The The first constraint states that there must be a station either in city 1 or in some
objective is to minimize the number of fire stations used. adjacent city. Notice that the constraint coefficient aij is 1 if city i is adjacent to city j or
if i=j and 0 otherwise.
The jth column of the constraint matrix represents the set of cities that can be served
by a fire station in city j. We are asked to find a set of such subsets j that covers the
set of all cities in the sense that every city appears in the service subset associated
with at least one fire station

78 79
5.1.5 Either - Or Constraints 5.1.6 If - Then Constraints
The following situation commonly occurs in mathematical programming problems. In many mathematical programming applications, the following situation occurs.
We are given two constraints of the form: We want to ensure that a constraint f(x1, x2,…, xn) > 0 implies the constraint g(x1,
f (x1, x2,…, xn) < 0 (1) x2,…, xn) > 0:
g (x1, x2,…, xn) < 0 (2) If f(x1, x2,…, xn) > 0 is satisfied, then the constraint g(x1, x2,…, xn) > 0 must be
We want to ensure that at least one of (1) and (2) is satisfied, often called either-or satisfied. While if f(x1, x2,…, xn) > 0 is not satisfied, then g(x1, x2,…, xn) > 0 may or
constraints. may not be satisfied.
Adding the two constraints (1’) and (2’) to the formulation will ensure our aim: To ensure this, we include the following constraints in the formulation:
f (x1, x2,…, xn) < M y (1’) – g(x1, x2,…, xn) < M y (1)
g (x1, x2,…, xn) < M ( 1 – y ) (2’) f(x1, x2,…, xn) < M ( 1 – y ) (2)
Here y is a 0-1 variable, and M is a number chosen large enough to ensure that f (x1, Here y is a 0-1 variable, and M is a large positive number.
x2,…, xn) < M and g (x1, x2,…, xn) < M are satisfied for all values of xj’s that satisfy the M must be chosen large enough so that – g < M and f < M hold for all values of xj’s
other constraints in the problem. that satisfy the other constraints in the problem.
Suppose y = 0, then (1) and possibly (2) must be satisfied. Observe that if f > 0, then (2) can be satisfied only if y = 0. (1) implies – g < 0 or g >
If y = 1, then (2) and possibly (1) must be satisfied. 0, which is the desired result.
Example Example
Suppose 1.5 tons of steel and 30 hours of labor are required for production of one For the lockbox problem, suppose we add the following constraint: If customers in
compact car. At present, 6.000 tons of steel and 60.000 hours of labor are available. region 1 send their payments to city 1, no other customers may send their payments
For an economically feasible production, at least 1000 cars of compact car must be to city 1. Mathematically,
produced. If x11 = 1, then x21 = x31 = x 41 = 0
Constraint: x1 < 0 or x1 > 1000 Since all variables are 0-1, we may write this implication as:
[f (x1, x2,…, xn) = x1; g (x1, x2,…, xn) = 1000 – x1] If x11 > 0, then x21 + x31 + x 41 < 0 (or – x21 – x31 – x 41 > 0)
Sign restriction: x1 > 0 and integer
If we define f = x11 and g = – x21 – x31 – x 41 we can use (1) and (2) to express the
We can replace this constraint by the following pair of linear constraints: implication by the following two constraints:
x1 < M1 y1 x21 + x31 + x 41 < My
1000 – x1 < M1 (1 – y1) x11 < M (1 – y)
y1 = 0 or 1 y = 0 or 1
M1 = min (6.000/1.5, 60.000/30) = 2000 Since –g and f can never exceed 3, we can choose M as 3.

80 81
5.1.7 Traveling Salesperson Problems 5.2 SOLVING INTEGER PROGRAMS
A salesperson must visit each of ten cities once before returning to his home. “What
ordering of the cities minimizes the total distance the salesperson must travel before We have gone through a number of examples of integer programs at the
returning home?” problem is called the traveling salesperson problem (TSP), not “Formulating IP Problems” section. A natural question is ``How can we get solutions
surprisingly. to these models?''. There are two common approaches. Historically, the first method
developed was based on cutting planes (adding constraints to force integrality). In
An IP Formulation of the TSP the last twenty years or so, however, the most effective technique has been based on
Suppose there are N cities. dividing the problem into a number of smaller problems in a tree search method
For i1j let cij = distance form city i to city j and called branch and bound. Recently (the last ten years or so), cutting planes have
Let cii = M (a very large number relative to actual distances) made a resurgence in the form of facets and polyhedral characterizations.
Also define xij as a 0-1 variable as follows: Actually, all these approaches involve solving a series of LP. For solving LP’s we
xij = 1 if the solution to TSP goes from city i to city j; have general purpose (independent of the LP being solved) and computationally
xij = 0 otherwise effective (able to solve large LP's) algorithms (simplex or interior point). For solving
The formulation of the TSP is: IP's no similar general purpose and computationally effective algorithms exist.
Min SUM[i] SUM[j] cij xij
st SUM[i] xij = 1 for all j Solution methods for IP's can be categorized as:
SUM[j] xij = 1 for all i ! General purpose (will solve any IP) but potentially computationally ineffective (will
ui – uj + N xij < N – 1 for i1j; i = 2, 3, …, N; j = 2, 3, …, N only solve relatively small problems); or
All xij = 0 or 1, All ui > 0 ! Special purpose (designed for one particular type of IP problem) but potentially
(SUM[j] should be read as "sum over all integer values of j from 1 to n inclusive") computationally more effective.
Solution methods for IP's can also be categorized as:
The first set of constraints ensures that s/he arrives once at each city. ! Optimal
The second set of constraints ensures that s/he leaves each city once. ! Heuristic
The third set of constraints ensure the following: An optimal algorithm is one which (mathematically) guarantees to find the optimal
Any set of xij’s containing a subtour will be infeasible solution.
Any set of xij’s that forms a tour will be feasible
It may be that we are not interested in the optimal solution:
REMARK ! because the size of problem that we want to solve is beyond the
The formulation of an IP whose solution will solve a TSP becomes unwieldy and computational limit of known optimal algorithms within the computer time we
inefficient for large TSPs. When using branch and bound methods to solve TSPs with have available; or
many cities, large amounts of computer time may be required. For this reason, ! we could solve optimally but feel that this is not worth the effort (time, money,
heuristics, which quickly lead to a good (but not necessarily optimal solution to a etc) we would expend in finding the optimal solution.
TSP, are often used.

82 83
In such cases we can use a heuristic algorithm - that is an algorithm that should 5.2.1 LP Relaxation (Relationship to LP)
hopefully find a feasible solution that, in objective function terms, is close to the For any IP we can generate an LP by taking the same objective function and same
optimal solution. In fact it is often the case that a well-designed heuristic algorithm constraints but with the requirement that variables are integer replaced by
can give good quality (near-optimal) results. appropriate continuous constraints:
Hence we have four categories that we potentially need to consider for solution “xi = 0 or 1” can be replaced by xi >= 0 and xi <= 1
algorithms: “xi > 0 and integer” can be replaced by xi >= 0
! General Purpose, Optimal The LP obtained by omitting all integer and 0-1 constraints on variables is called the
Enumeration, branch and bound, cutting plane LP Relaxation of the IP. We can then solve this linear relaxation (LR) of the original
! General Purpose, Heuristic IP.
Running a general purpose optimal algorithm and terminating after a specified If LR is optimized by integer variables (solution is turned out to have all variables
time taking integer values at the optimal solution), then that solution is feasible and
! Special Purpose, Optimal (beyond the scope of this course) optimal for IP (naturally integer LP)
Tree search approaches based upon generating bounds via dual ascent, Since LR is less constrained than IP, the following are immediate:
lagrangean relaxation ! If IP is a maximization, the optimal objective value for LR is greater than or
! Special Purpose, Heuristic (beyond the scope of this course) equal to that of IP.
Bound based heuristics, tabu search, simulated annealing, population ! If IP is a minimization, the optimal objective value for LR is less than or equal
heuristics (e.g. genetic algorithms), interchange to the optimal objective for IP.
! If LR is infeasible, then so is IP.
So solving LR does give some information: it gives a bound on the optimal value,
and, if we are lucky, may give the optimal solution to IP. We saw, however, that
rounding the solution of LR will not in general give the optimal solution of IP. In fact,
for some problems it is difficult to round and even get a feasible solution.
In general the solution process used in many LP based IP packages is
! specify the IP: objective, constraints, integer variables (typically xj = integer
between 0 and n).
! automatically generate the LP relaxation: same objective as IP, same
constraints as IP with the addition of 0<=xj<=n, variables as before but no
longer required to be integer
! use the branch and bound to generate the optimal solution to the IP

84 85
5.2.2 Enumeration Hence for our example we merely have to examine 16 possibilities before we know
Unlike LP (where variables took continuous values (>=0)) in IP's (where all variables precisely what the best possible solution is. This example illustrates a general truth
are integers) each variable can only take a finite number of discrete (integer) values. about integer programming:
Hence the obvious solution approach is simply to enumerate all these possibilities - What makes solving the problem easy when it is small is precisely what makes it
calculating the value of the objective function at each one and choosing the (feasible) hard very quickly as the problem size increases
one with the optimal value. This is simply illustrated: suppose we have 100 integer variables each with two
For example for the multi-period capital budgeting problem, possible integer values then there are 2x2x2x ... x2 = 2100 (approximately 1030)
Maximize 0.2 x1 + 0.3 x2 + 0.5 x3 + 0.1 x4 possibilities which we have to enumerate (obviously many of these possibilities will
Subject to 0.5 x1 + 1 x2 + 1.5 x3 + 0.1 x4 < 3.1
be infeasible, but until we generate one we cannot check it against the constraints to
0.3 x1 + 0.8 x2 + 1.5 x3 + 0.4 x4 < 2.5
see if it is feasible).
0.2 x1 + 0.2 x2 + 0.3 x3 + 0.1 x4 < 0.4
xj = 0 or 1 j = 1, … 4
Be clear here - conceptually there is not a problem - simply enumerate all
4
there are 2 =16 possible solutions. These are: possibilities and choose the best one. But computationally (numerically) this is just
x1 x2 x3 x4 impossible.
0 0 0 0 do no projects IP nowadays is often called "combinatorial optimization" indicating that we are
0 0 0 1 do one project
dealing with optimization problems with an extremely large (combinatorial) increase
0 0 1 0
in the number of possible solutions as the problem size increases.
0 1 0 0
1 0 0 0
0 0 1 1 do two projects
0 1 0 1
1 0 0 1
0 1 1 0
1 0 1 0
1 1 0 0
1 1 1 0 do three projects
1 1 0 1
1 0 1 1
0 1 1 1
1 1 1 1 do four projects

86 87
5.2.3 Branch and Bound Consider this LP relaxation solution. We have a variable x2 which is fractional when
The most effective general purpose optimal algorithm is LP-based tree search (also we need it to be integer. To resolve this we can generate two new problems:
widely being called as branch and bound). P1: original LP relaxation plus x2=0
This is a way of systematically enumerating feasible solutions such that the optimal P2: original LP relaxation plus x2=1
integer solution is found. Then we will claim that the optimal integer solution to the original problem is
Where this method differs from the enumeration method is that not all the feasible contained in one of these two new problems. This process of taking a fractional
solutions are enumerated but only a fraction (hopefully a small fraction) of them. variable and explicitly constraining it to each of its integer values is known as
However we can still guarantee that we will find the optimal integer solution. The branching. It can be represented diagrammatically as below (in a tree diagram,
method was first put forward in the early 1960's by Land and Doig. which is how the name tree search arises).
5.2.3.1 Example
Consider our example multi-period capital budgeting problem:
Maximize 0.2 x1 + 0.3 x2 + 0.5 x3 + 0.1 x4
Subject to 0.5 x1 + 1 x2 + 1.5 x3 + 0.1 x4 < 3.1
0.3 x1 + 0.8 x2 + 1.5 x3 + 0.4 x4 < 2.5
0.2 x1 + 0.2 x2 + 0.3 x3 + 0.1 x4 < 0.4
xj = 0 or 1 j = 1, … 4 We now have two new LP relaxations to solve. If we do this we get:
What made this problem difficult was the fact that the variables were restricted to be P1 solution is x1=0.5, x3=1, x2=x4=0 of value 0.6
integers (zero or one). P2 solution is x2=1, x3=0.67, x1=x4=0 of value 0.63
If the variables had been allowed to be fractional (takes all values between zero and This can be represented diagrammatically as below.
one for example) then we would have had an LP which we could easily solve.
Suppose that we were to solve this LP relaxation of the problem [replace xj = 0 or 1
j=1,...,4 by 0 <= xj <= 1 j=1,...,4]. Then using any LP package we get x2=0.5, x3=1,
x1=x4=0 of value 0.65 (i.e. the objective function value of the optimal linear
programming solution is 0.65).
As a result of this we now know something about the optimal integer solution, namely
that it is <= 0.65, i.e. this value of 0.65 is an upper bound on the optimal integer
solution. To find the optimal integer solution we just repeat the process, choosing one of these
This is because when we relax the integrality constraint we (as we are maximizing) two problems, choosing one fractional variable and generating two new problems to
end up with a solution value at least that of the optimal integer solution (and maybe solve.
better).

88 89
Choosing problem P1 we branch on x1 to get our list of LP relaxations as: Branching on x3 we get
P3 (P1 plus x1=0) solution x3=x4=1, x1=x2=0 of value 0.6 P5 (P2 plus x3=0) solution x1=x2=1, x3=x4=0 of value 0.5
P4 (P1 plus x1=1) solution x1=1, x3=0.67, x2=x4=0 of value 0.53 P6 (P2 plus x3=1) solution infeasible
P2 solution x2=1, x3=0.67, x1=x4=0 of value 0.63 Neither of P5 or P6 lead to further branching so we are done, we have discovered
This can again be represented diagrammatically as below. the optimal integer solution of value 0.6 corresponding to x3=x4=1, x1=x2=0.
The entire process we have gone through to discover this optimal solution (and to
prove that it is optimal) is shown graphically below.
At this stage we have identified an integer feasible solution of value 0.6 at P3. There
are no fractional variables so no branching is necessary and P3 can be dropped from
our list of LP relaxations.
Note here that this method, like complete enumeration, also involves powers of two
Hence we now have new information about our optimal (best) integer solution,
as we progress down the (binary) tree. However also note that we did not enumerate
namely that it lies between 0.6 and 0.65 (inclusive).
all possible integer solutions (of which there are 16). Instead here we solved 7 LP's.
Consider P4, it has value 0.53 and has a fractional variable (x3). However if we were
This is an important point, and indeed why tree search works at all. We do not need
to branch on x3 any objective function solution values we get after branching can
to examine as many LP's as there are possible solutions. While the computational
never be better (higher) than 0.53. As we already have an integer feasible solution of
efficiency of tree search differs for different problems it is this basic fact that enables
value 0.6, P4 can be dropped from our list of LP relaxations since branching from it
us to solve problems that would be completely beyond us were we to try complete
could never find an improved feasible solution. This is known as bounding - using a
enumeration.
known feasible solution (as a lower bound) to identify that some relaxations are not
of any interest and can be discarded.
Hence we are just left with:
P2 solution x2=1, x3=0.67, x1=x4=0 of value 0.63

90 91
5.2.3.2 An Example of Branch and Bound in Graphical Solution 5.2.4 The Cutting Plane Algorithm
(Winston 9.3., p. 503) There is an alternative to branch and bound called cutting planes which can also be
LP Relaxation and the first two subproblems used to solve integer programs.
(http://www.isl.itu.edu.tr/ya/branch_and_bound_graphical_f1.jpg) The fundamental idea behind cutting planes is to add constraints to a linear program
Rest of the solution until the optimal basic feasible solution takes on integer values. Of course, we have
(http://www.isl.itu.edu.tr/ya/branch_and_bound_graphical_f2.jpg) to be careful which constraints we add: we would not want to change the problem by
adding the constraints. We will add a special type of constraint called a cut.
LINDO OUTPUT: A cut relative to a current fractional solution satisfies the following criteria:
MAX 8 X1 + 5 X2 Every feasible integer solution is feasible for the cut, and
SUBJECT TO
2) X1 + X2 <= 6 The current fractional solution is not feasible for the cut.
3) 9 X1 + 5 X2 <= 45
This is illustrated in the figure given below.
END
GIN 2
LP OPTIMUM FOUND AT STEP 2

OBJECTIVE VALUE = 41.2500000
SET X1 TO <= 3 AT 1, BND= 39.00 TWIN= 41.00 9
NEW INTEGER SOLUTION OF 39.0000000 AT BRANCH 1 PIVOT 9

BOUND ON OPTIMUM: 41.00000
FLIP X1 TO >= 4 AT 1 WITH BND= 41.000000
SET X2 TO <= 1 AT 2, BND= 40.56 TWIN=-0.1000E+31 12
SET X1 TO >= 5 AT 3, BND= 40.00 TWIN= 37.00 15
NEW INTEGER SOLUTION OF 40.0000000 AT BRANCH 3 PIVOT 15

BOUND ON OPTIMUM: 40.00000
DELETE X1 AT LEVEL 3
ENUMERATION COMPLETE. BRANCHES= 3 PIVOTS= 15
There are two ways to generate cuts:
LAST INTEGER SOLUTION IS THE BEST FOUND
RE-INSTALLING BEST SOLUTION... Gomory cuts generates cuts from any linear programming tableau. This has the
advantage of “solving'' any problem but has the disadvantage that the method can be
OBJECTIVE FUNCTION VALUE
very slow.
1) 40.00000
The second approach is to use the structure of the problem to generate very good
VARIABLE VALUE REDUCED COST
cuts. The approach needs a problem-by-problem analysis, but can provide very
X1 5.000000 -8.000000
X2 0.000000 -5.000000 efficient solution techniques.
ROW SLACK OR SURPLUS DUAL PRICES
2) 1.000000 0.000000
3) 0.000000 0.000000
NO. ITERATIONS= 15
BRANCHES= 3 DETERM.= 1.000E 0

92 93
5.2.4.1 Steps 5.2.4.3 Example
1. Find the optimal tableau for the IP’s LP relaxation. If all variables in the optimal Consider the following integer program:
solution assume integer values, we have found an optimal solution! Otherwise max z = 8 x1 + 5 x2
st x1 + x2 < 6
proceed to next step
9x1 + 5x2 < 45
2. Pick a constraint in the optimal tableau whose RHS has the fractional part closest x1, x2 > 0 and integer
to ½. If we ignore integrality, we get the following optimal tableau:
3. For the constraint identified, put all of the integer parts on the left side (round z x1 x2 s1 s2 RHS
1 0 0 1.25 0.75 41.25
down), and all the fractional parts on the right
0 0 1 2.25 - 0.25 2.25
4. Generate the cut as: “RHS of the modified constraint” < 0 0 1 0 - 1.25 0.25 3.75
5. Use the dual simplex to find the optimal solution to the LP relaxation, with the cut Let's choose the constraint whose RHS has the fractional part closest to ½::
as an additional constraint. If all variables in the optimal solution assume integer x1 – 1.25 s1 + 0.25 s2 =3.75 (Arbitrarily choose the second constraint)
values, we have found an optimal solution. Otherwise return to Step 2. We can manipulate this to put all of the integer parts on the left side (round down),
and all the fractional parts on the right to get:
5.2.4.2 Dual Simplex Method (for a Max Problem) x1 – 2 s1 + 0 s2 – 3 = 0.75 – 0.75 s1 – 0.25 s2
We choose the most negative RHS. Now, note that the left hand side consists only of integers, so the right hand side
BV of this pivot row leaves the basis. must add up to an integer. It consists of some positive fraction minus a series of
For the variables that have a negative coefficient in the pivot row, we calculate the positive values. Therefore, the right hand side cannot be a positive value. Therefore,
ratios (coefficient in R0 / coefficient in pivot row). we have derived the following constraint:
Variable with the smallest ratio (absolute value) enters basis. 0.75 – 0.75 s1 – 0.25 s2 < 0
This constraint is satisfied by every feasible integer solution to our original problem.
But, in our current solution, s1 and s2 both equal 0, which is infeasible to the above
constraint. This means the above constraint is a cut, called the Gomory cut after its
discoverer.
We can now add this constraint to the linear program and be guaranteed to find a
different solution, one that might be integer.
z x1 x2 s1 s2 s3 RHS
1 0 0 1.25 0.75 0 41.25
0 0 1 2.25 - 0.25 0 2.25
0 1 0 - 1.25 0.25 0 3.75
0 0 0 - 0.75 - 0.25 1 - 0.75
The dual simplex ratio test indicates that s1 should enter the basis instead of s3. The
optimal solution is an IP solution:
z = 40, x1 = 5, x2 = 0

94 95

Operations Research PDF

Uploaded by

Copyright:

Available Formats

Operations Research PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Operations Research PDF

Uploaded by

Copyright:

Available Formats

What are the basic concepts of Operations Research discussed?

What are the basic concepts of Operations Research discussed?

What are some examples used to illustrate linear programming formulations?

What are some examples used to illustrate linear programming formulations?

CONTENTS

Dr. Y. !lker Topcu (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

4.2 FINDING BFS FOR TRANSPORT’N PROBLEMS .......................................... 50 "ORMS".

4.2.2 Minimum Cost Method.................................................................................. 53 science" ("DS"), and “problem solving”.

4.3.1 Steps of the Method ...................................................................................... 57

4.4.1 Steps................................................................................................................ 61 procedure should be followed:

4.4.2 Kuruoglu Example ......................................................................................... 62 Step 1. Formulate the Problem

4.5.3 Flight Crew Example ..................................................................................... 66 Step 2. Observe the System

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

formulating the model). model to be a linear program (LP) were:

the formulation of the problem. ! a single objective (minimize or maximize)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

180. (Optimal solution) x1 = the number of comedy spots

buying more labor. st 7x1 + 2x2 $% 28

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

We determine the optimal solution to the LP by plotting 180x + 160y = K (K constant)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

4. Find a new bfs

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

The new pivot equation (R3/2) is THE CURRENT SOLUTION IS OPTIMAL

R3’ : x1+.75x2+.25x3+ .5s3 =4

R0’ z +15x2 -5x3 +30s3 = 240 z = 240 (R0+60R3’

The new bfs is x2 = x3 = s3 = 0, x1 = 4, s1 = 16, s2 = 4, s4 = 5 making z = 240.

5. Go back to Step 3. Repeat steps until an optimal solution is reached

Rearranging the pivot equation gives

R2’’ -2x2+x3+2s2-4s3 = 8 (R2’×2).

Row operations with R2’’ eliminate x3 to give the new system

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

4 2 1.5 0 1 0 0 20 s2 = 20 1 0 0 0 0 10 10 0 280 z=280

0 1 1.25 0 0 -.5 1.5 0 2 x1=2 2/1.25 )

Second and optimal tableau: 0 -0.8 0 0 0 0.4 -1.2 1 3.4 s4=3.4

1 1.25 0 0 -.5 1.5 0 2 x1 = 2 x3 8 11.2 11.2 – 3.2c

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

2. Convert each inequality constraint to standard form

0.5 x1+ 0.25 x2 + s1 = 4 (

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

3.3.2 Shadow Price

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

NOTICE: The objective function which is regarded as Row 0 in Simplex is

ROW SLACK OR SURPLUS DUAL PRICES

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)

4x1 + 2x2 + 1.5x3 +s2 = 20 Finishing

2x1 + 1.5x2 + .5x3 +s3 = 8 Carpentry

problem): 8x1' + 6x2' + x3' +s1' = 48

2x1' + 1.5x2' + .5x3' +s3' = 8

Y. !lker Topcu, Ph.D. (www.ilkertopcu.net) Y. !lker Topcu, Ph.D. (www.ilkertopcu.net)