Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Subject Code: MB
0048
BKID B1631
Additional Registrar
SMU DDE
Prof. K. V. Varambally
Director
Manipal Institute of Management, Manipal
Originally Prepared by
Prof. S. Santhanam
Faculty
MP Birla Institute of Management
Contents
Unit 1
Introduction to Operations Research
Unit 2
Linear Programming
21
Unit 3
Graphical Analysis of Linear Programming Problems
39
Unit 4
Simplex Method
57
Unit 5
Duality in Linear Programming Problem
87
Unit 6
Transportation Problem
109
Unit 7
Assignment Problem
132
Unit 8
Integer Programming Problem
154
Unit 9
Infinite Queuing Models
174
Unit 10
Mathematical Analysis of Queuing Theory
187
Unit 11
Finite Queuing Models
213
Unit 12
Simulation
227
Unit 13
Simulation Monte-Carlo Method
252
Unit 14
Project Scheduling and PERT-CPM
274
Unit 15
Game Theory
295
MB 0048
Operations Research
Course Description
Operations Research is a systematic and scientific study of the operations
of the system. It provides executives and managers a quantitative basis for
taking decisions regarding the operations under their control.
Operations Research came into existence in the military context during
World War II. It was used to solve strategic and technical problems to
improve the execution of various military projects. Following the end of war,
this new science attracted the attention of industrial managers who were
seeking solutions to complex industrial problems. Thus they started using
this to solve problems and to make decisions.
Today, in many countries, almost every large organisation or corporation
uses this subject for solving complex problems and making decisions.
Managers of many organisations have understood the advantages of the
scientific approach on which operations research is based.
Operations Research is currently a popular subject in almost every
management institute. This Operations Research course is a four credit
course in MBA second semester. It introduces students to many OR
concepts and elaborates various topics on linear programming, queuing
theory, assignment problems, PERT and CPM, etc.
Course Objectives:
After studying this course, the student will be able to:
explain the scope, importance and features of operations research
define linear programming
solve linear programming problem
solve linear programming problem (LPP) using graphical method
explain the concept of duality in LPP and solve the dual problem
solve the transportation problem of minimising the cost of transporting a
commodity
The Self Learning Material (SLM) for this subject is divided into 15 units. A
brief description of all the 15 units is given below:
Unit 1: Introduction to Operations Research
This unit defines operations research. It explains the scope, features,
phases, research methodology, techniques, tools, and limitations of
operations research.
Unit 2: Linear Programming
This unit describes the linear programming problem (LPP), its mathematical
formulation, application, advantages and disadvantages, and its basic
assumptions.
Unit 3: Graphical Analysis of Linear Programming Problems
This unit deals with graphical methods to solve LPP, basic definitions,
working rule, classification of feasible points, maximisation of LPP,
exceptional cases and important properties of LPP.
Unit 4: Simplex Method
This unit explains the standard form of LPP, solution of LPP using simplex
method, simplex algorithm, two phase method and penalty cost method.
Unit 5: Duality in LPP
This unit describes the importance of duality concepts, formulation of dual
concepts, economic interpretation of duality and sensitivity analysis.
Unit 6: Transportation Problem
This unit describes the formulation of Transportation Problem (TP), the initial
basic solution, and transportation algorithm. It also explains how to get the
optimum solution for TP using U-V method.
Operations Research
Unit 1
Unit 1
Structure:
1.1
Introduction
Objectives
1.2
Historical Background
Definitions of operations research
1.3
Scope of Operations Research
1.4
Features of Operations Research
1.5
Phases of Operations Research
1.6
Types of Operations Research Models
A broad classification of OR models
1.7
Operations Research Methodology
Definition
Construction
Solution
Validation
Implementation
1.8
Operations Research Techniques and Tools
1.9
Structure of the Mathematical Model
1.10 Limitations of Operations Research
1.11 Summary
1.12 Glossary
1.13 Terminal Questions
1.14 Answers
1.15 Case Study
1.1 Introduction
Welcome to the unit on operations research management. Operations
research management focuses on the mathematical scoring of
consequences of a decision aiming to optimise the use of time, effort, and
resources to avoid blunders. The act of obtaining best results under any
given circumstances is known as optimising. The key purpose of Operations
Research (OR) is to do preparative calculations that aid the decision-making
process.
Sikkim Manipal University
Page No. 1
Operations Research
Unit 1
Hence, decision-making is a key part of our daily life. The ultimate goal of all
decisions is to maximise benefits and to minimise effort and time. OR gives
decision makers the power to make effective decisions and improve day-today operations. Decision makers consider all the available options, study
the outcomes, and estimate the risks.
In simple situations, common sense and judgement can be used to take
decisions. For example, if you are buying a microwave or washing machine,
the decision-making process is not very complicated. You can simply
compare the price, quality, and durability of the well-known brands and
models in the market and take a decision based on it.
However, in complex situations, although it is possible to take decisions
based on ones common sense, a decision backed by mathematical
calculations reduces the risk factor and increases the probability of success.
Some such situations, where decision-makers have to depend on
mathematical scoring and reasoning, are finding an appropriate product mix
amidst competitors products or planning a public transportation network in a
city.
Objectives:
After studying this unit, you should be able to:
describe the historical background of OR
list the significant features of OR
describe the methodology of OR
define the structure of a mathematical model in OR
describe the significance of the function of OR
Page No. 2
Operations Research
Unit 1
Page No. 3
Operations Research
Unit 1
Page No. 4
Operations Research
Unit 1
Page No. 5
Operations Research
Unit 1
Page No. 6
Operations Research
Unit 1
Action phase
This phase involves making recommendations for the decision process. The
recommendations can be made by those who identify and present the
problem or by anyone who influences the operation in which the problem
has occurred.
Self Assessment Questions
7. Action phase involves making recommendations for the decision
process. (True/False)
8. One of the OR phases is judgement phase. (True/False)
Page No. 7
Operations Research
Unit 1
Probabilistic models - These are the models in which the input and
output variables follow a defined probability distribution, such as the
games theory.
General models These are the models which you can apply in general
to any problem. For example, linear programming.
Specific models - These are the models that you can apply only under
specific conditions. For example, you can use the sales response curve
or equation as a function in the marketing function.
Page No. 8
Operations Research
Unit 1
As shown in figure 1.3, OR methodology consists of five steps. They are defining the problem, constructing the model, solving the model, validating
the model, and implementing the result.
Let us now study the steps in detail.
1.7.1 Definition
The first and the most important step in the OR approach of problem solving
is to define the problem. One needs to ensure that the problem is identified
properly because this problem statement will indicate the following three
major aspects:
Description of the goal or the objective of the study
Identification of the decision alternative to the system
Sikkim Manipal University
Page No. 9
Operations Research
Unit 1
1.7.2 Construction
Based on the problem definition, you need to identify and select the most
appropriate model to represent the system. While selecting a model, you
need to ensure that the model specifies quantitative expressions for the
objective and the constraints of the problem in terms of its decision
variables. A model gives a perspective picture of the whole problem and
helps in tackling it in a well-organised manner. Therefore, if the resulting
model fits into one of the common mathematical models, you can obtain a
convenient solution by using mathematical techniques. If the mathematical
relationships of the model are too complex to allow analytic solutions, a
simulation model may be more appropriate. Hence, appropriate models can
be constructed.
1.7.3 Solution
After deciding on an appropriate model, you need to develop a solution for
the model and interpret the solution in the context of the given problem. A
solution to a model implies determination of a specific set of decision
variables that would yield an optimum solution. An optimum solution is one
which maximises or minimises the performance of any measure in a model
subject to the conditions and constraints imposed on the model.
1.7.4 Validation
A model is a good representation of a system. However, the optimal solution
must work towards improving the systems performance. You can test the
validity of a model by comparing its performance with some past data
available from the actual system. If under similar conditions of inputs, your
model can reproduce the past performance of the system, then you can be
sure that your model is valid. However, you will still have no assurance that
future performance will continue to duplicate the past behaviour. Secondly,
since the model is based on careful examination of past data, the
comparison should always reveal favourable results. In some instances, this
problem may be overcome by using data from trial runs of the system. One
must note that such validation methods are not appropriate for non-existent
systems because data will not be available for comparison.
Page No. 10
Operations Research
Unit 1
1.7.5 Implementation
You need to apply the optimal solution obtained from the model to the
system and note the improvement in the performance of the system. You
need to validate this performance check under changing conditions. To do
so, you need to translate these results into detailed operating instructions
issued in an understandable form to the individuals who will administer and
operate the recommended system. The interaction between the operations
research team and the operating personnel reaches its peak in this phase.
Page No. 11
Operations Research
Unit 1
i 1
ai
bj
(1)
j 1
In addition to the above, there are tools such as the sequence model,
the assignment model, and network analysis, which you will learn in
detail in the later units.
Page No. 12
Operations Research
Unit 1
Page No. 13
Operations Research
Unit 1
Page No. 14
Operations Research
Unit 1
Page No. 15
Operations Research
Unit 1
Since the number of units (x1 or x2) are always non-negative, you have x1
0, x2 0.
Therefore, the mathematical model is:
5x1 + 6x2 60
2x1 + 3x2 80
x1 0, x2 0, min z = 5x1 + 6x2.
Time and money factors When you subject the basic data to frequent
changes then incorporating them into OR models becomes a costly
affair.
Page No. 16
Operations Research
Unit 1
1.11 Summary
Let us recapitulate the important concepts discussed in this unit:
The objective of OR is to provide a scientific basis to the decisionmakers for solving problems involving interaction with various
components of the organisation
1.12 Glossary
Probability: possible outcomes of an event
Hypothesis: unproved theory
Network analysis: a mathematical representation of the problem by lines
and nodes joined to form a network
Page No. 17
Operations Research
Unit 1
Define OR.
What are the features of OR?
What is a model in OR? Discuss the different models available in OR.
Write short notes on the different phases of OR.
What are the limitations of OR?
1.14 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Scientific basis
Scientists, different disciplines
economic , social policies
Outpatient, administrative
Imbibes
Decision making
True
True
True
False
False
False
Inter-disciplinary
OR
Judgement phase, research phase, and action phase
Define the problem
Model
Parameters
Terminal Questions
1. OR is defined as the application of scientific methods, techniques and
tools to the operation of a system with optimum solutions to the
problems - refer 1.2.1
2. Some key features of OR are OR is system oriented, it imbibes inter
disciplinary team approach and uses scientific methods to solve
problems - refer 1.4
3. Types of operations research models - refer 1.6
Sikkim Manipal University
Page No. 18
Operations Research
Unit 1
Page No. 19
Operations Research
Unit 1
Kanti Swarup & Gupta P. K., & Hira D. S., & Manmohan (2004).
Operation Research. Sultan Chand and Sons.
Cohen M., & Kamesan P. V., & Kleindorfer P., & Lee H., & Tekerian A.
(Jan-Feb. 1990). Optimizer: IBMs Multi-Echelon Invenentory System
for Managing Service Logistics.
E- References:
newagepublishers.com.
http://www.newagepublishers.com/samplechapter/001012.pdf.
Page No. 20
Operations Research
Unit 2
Unit 2
Linear Programming
Structure:
2.1 Introduction
Objectives
2.2 Linear Programming Problem
2.3 Requirements of LPP
Basic assumptions of LPP
2.4 Mathematical Formulation of LPP
General form of LPP
Canonical forms
2.5 Case Studies of LPP
Solved problem
2.6 Applications of LPP
2.7 Advantages of LPP
2.8 Limitations of LPP
2.9 Summary
2.10 Glossary
2.11 Terminal Questions
2.12 Answers
2.13 Case Study
2.1 Introduction
In the previous unit, we dealt with operations research. We discussed the
background, scope, features, phases, types, limitations, techniques, and
tools of operations research. We also analysed the operations research
methodology and the structure of the mathematical model. In this unit, we
will deal with linear programming. Linear programming focuses on obtaining
the best possible output (or a set of outputs) from a given set of limited
resources.
Minimal time and effort and maximum benefit coupled with the best possible
output or a set of outputs is the mantra of any decision-maker. Today,
decision-makers or managements have to tackle the issue of allocating
limited and scarce resources at various levels in an organisation, in the best
possible manner. Man, money, machine, time, and technology are some of
Page No. 21
Operations Research
Unit 2
Page No. 22
Operations Research
Unit 2
Page No. 23
Operations Research
Unit 2
Solved problem 1:
A company has three operational departments (weaving, processing, and
packing) with a capacity to produce three different types of clothes - suits,
shirts, and woollens, yielding a profit of Rs.2, Rs.4, and Rs.3 per meter
respectively. One meter of suiting requires 3 minutes of weaving, 2 minutes
of processing, and 1 minute of packing. Similarly, one meter of shirting
requires 4 minutes of weaving, 1 minute of processing, and 3 minutes of
packing. One meter of woollen requires 3 minutes in each department. In a
week, total run time of each department is 60, 40, and 80 hours for weaving,
processing, and packing respectively.
Formulate the linear programming problem to find the product mix to
maximise the profit.
Page No. 24
Operations Research
Unit 2
Mathematical formulation:
The data of the problem is summarised below in table 2.1:
Table 2.1: summary of the data of the problem
Departments
Profit
Weaving
(in min)
Processing
(in min)
Packing
(in min)
(Rs. per
meter)
Suits
Shirts
Woollens
60*60
40*60
80*60
Availability (min)
Page No. 25
Operations Research
Unit 2
The linear programming problem can thus be put in the following algebraic
format:
Find x1, x2, and x3 so as to maximise
Z = 2x1+4x2+3x3
Subject to constraints:
3x1+4x2+3x3 3600
2x1+x2+3x3 2400
3x1+3x2+3x3 4800
x1, x2, and x3 0
Page No. 26
Operations Research
Unit 2
Page No. 27
Operations Research
Unit 2
4. An inequality constraint with its left hand side in the absolute form can
be changed into two regular inequalities. For example,
|2x1+3x2| 5 is equivalent to 2x1+3x2 5 and 2x1+3x2 5 or 2x1 3x2 5.
Assembly
Painting
Testing
1.0
1.5
0.2
0.2
0.0
0.1
Total numbers of hours available each week are as under assembly 600,
painting 100, and testing 30. The firm wishes to determine the weekly
product-mix to maximise revenue.
Sikkim Manipal University
Page No. 28
Operations Research
Unit 2
Total revenue
x1:
x2:
x1, x2:
Decision variables
Since the objective (goal) of the firm is to maximise its revenue, the
model can be stated as follows. The objective function Z = 50x1 + 80x2 is
to be maximised subject to the following constraints:
1.0x1 + 1.5x2 600, (assembly constraints)
0.2x1+0.2x2 100, (painting constraints)
0.0x1+0.1x2 30, (testing constraints)
x10, x20, (the non-negativity constraints)
Solved problem 2
A milk distributor supplies milk in bottles to houses in three areas A, B,
and C in a city. His delivery charge per bottle is 30 paise in area A, 40
paise in area B, and 50 paise in area C. He has to spend on an average,
1 minute to supply one bottle in area A, 2 minutes per bottle in area B,
and 3 minutes per bottle in area C. He can spare only 2 hours 30 minutes
for milk distribution but not more than 1 hour 30 minutes for area A and B
together. He can deliver at the most 120 bottles.
Find the number of bottles that he has to supply in each area so as to
maximise the supply. Construct a mathematical model.
Page No. 29
Operations Research
Unit 2
Solved problem 3
An oil company has two units A and B, which produce three different
grades of oil super fine, medium, and low grade. The company has to
supply 12, 8, and 24 barrels of super fine, medium, and low-grade oil
respectively per week. It costs the company Rs.1000 and Rs.800 per day
to run the units A and B respectively.
In a day, unit A produces 6, 2, and 4 barrels and unit B produces 2, 2,
and 12 barrels of super fine, medium, and low grade respectively. The
manager has to decide on how many days per week should each unit be
operated in order to meet the requirement at minimum cost.
Formulate the LPP model.
Solution: Table 2.2 depicts the summary of the given data.
Table 2.2: Capacity and Requirements Details of an Oil Company
Product
Capacity
Super fine
Medium
Low grade
Cost
Requirements
Unit A
Unit B
6
2
4
Rs. 1000
2
2
12
Rs. 800
12
8
24
-
Let x1 and x2 be the number of days that unit A and B must be operated
per week respectively. Then the objective of the manager is to:
Minimise the cost function
Z = 1000 x1 + 800 x2
Subject to the constraints 6x1 + 2x2 12 (super fine)
2x1 + 2 x2 8 (medium)
4x1 + 12 x2 24 (low grade)
and x1 >0, x2 0
2.5.1 Solved problem
A toy company manufactures two types of doll, a basic version-doll A and a
deluxe version-doll B. Each doll of type B takes twice as long to produce
one doll of type A. Thus, the company would have time to make a maximum
Sikkim Manipal University
Page No. 30
Operations Research
Unit 2
p= 3x1+5x2
Subject to restrictions
x1+2x2 2000(time constraint)
x1+x2 1500(plastic constraint)
x2 600 (dress constraint)
and non-negativity restrictions
x1 0, x20
Self Assessment Questions
4. One of the characteristics of canonical form in the objective function
must be maximisation. (True/False)
5. 2x 3y 10 can be written as -2x + 3y -10. (True/False)
Page No. 31
Operations Research
Unit 2
Page No. 32
Operations Research
Unit 2
Page No. 33
Operations Research
Unit 2
Page No. 34
Operations Research
Unit 2
2.9 Summary
Let us recapitulate the important concepts discussed in this unit:
In LPP, first identify the decision variables with economic or physical
quantities whose values are of interest to the management.
The problems must have a well-defined objective function expressed in
terms of the decision variable.
The objective is to maximise the resources when it expresses profit or
contribution. If the objective function indicates cost, then it has to be
minimised.
The decision variables interact with each other through some
constraints. These constraints arise due to limited resources, stipulation
on quality, technical, legal, or a variety of other reasons.
2.10 Glossary
Decision variables - set of unknown quantities in LPP
Bottleneck - A point in the production process that causes delay.
Page No. 35
Operations Research
Unit 2
2.12 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
9.
Linear
Alternate course of action
Fractions
True
True
Linear programming
Decision problem
Proper mix of media
Raw material
Page No. 36
Operations Research
Unit 2
Terminal Questions
1. LPP is a mathematical technique for planning and decision making.
Examples: development of production schedule and establishment of an
investment portfolio. For more information refer 2.2
2. Applications of LPP in the area of finance, productions and operations
management, human resources, marketing and distributions. For more
information refer 2.6
3. Advantages of LPP: optimum utilisation of productive resources,
improves quality of decisions, provides practically applicable solutions.
For more information refer 2.7
4. Basic assumptions of LPP: Linearity, deterministic, additivity, divisibility
etc. For more information refer 2.3 and 2.3.1
5. The general linear programming model can be defined as follows:
Maximise or minimise Z = c1 x1 + c2 x2 + - - - - + cn xn
For more information refer 2.4.1
6. [Hint: x1= tons of grade A coal, x2= tons of grade B coal & x3= tons of
grade C coal]. For more information refer 2.4 and 2.5.
7. Maximise 3x1 + 5x2, subject to x1 + 2x2 800 (minutes)
3x1 + 2x2 1200 (minutes) x1, x2 0 For more information refer 2.4
and 2.5.
Page No. 37
Operations Research
Unit 2
purchased. The expected net annual profit would be Rs. 15,000 for the
station wagon, Rs. 35,000 for the mini bus, and Rs. 45,000 for the large bus.
The company has hired 30 new drivers for the new vehicles. They are
qualified drivers for all the three types of vehicles.
The maintenance department has the capacity to handle an additional 80
station wagons. A mini bus is equivalent to 5/3 station wagons and each
large bus is equivalent to two station wagons in terms of their use of the
maintenance department.
Discussion Questions:
1. Formulate a linear programming model, which would help to determine
the optimum number of each type of vehicle to be purchased in order to
maximise profit.
[Hint: x1= no. of station wagons to purchase, x2= no. of minibus to
purchase & x3= no. of large buses to purchase].
References :
Kanti Swarup & Gupta P. K., & Hira D. S., & Manmohan (2004). Operation
Research. Sultan Chand and Sons.
E-References:
Linear Optimization.
http://www.mirrorservice.org/sites/home.ubalt.edu/ntsbarsh/Businessstat/opre/partVIII.htm.
Page No. 38
Operations Research
Unit 3
Unit 3
Graphical Analysis of
Linear Programming Problems
Structure:
3.1 Introduction
Objectives
3.2 Graphical Analysis
3.3 Some Basic Definitions
3.4 Graphical Methods to Solve LPP
Working rule
Classification of feasible points
Example on maximisation of LPP
Solved problems on mixed constraints LP problem
3.5 Some Exceptional Cases
Solved problem for alternative optimal solution
Solved problem for unbounded solution
Solved problem for inconsistent solution
3.6 Important Geometric Properties of LPP
3.7 Summary
3.8 Glossary
3.9 Terminal Questions
3.10 Answers
3.11 Case Study
3.1 Introduction
In the previous unit, we dealt with Linear Programming Problem (LPP). We
analysed the definition, requirements, mathematical formulation, case
studies, applications, advantages, and disadvantages of LPP. In this unit,
we will deal with the graphical analysis of LPP. Today, many resources
needed as inputs to operations are in limited supply. Operations managers
must understand the impact of this situation on meeting their objectives.
Linear Programming (LP) is one way through which the operations
managers can determine how best to allocate their scarce resources.
Once the linear programming model has been formulated on the basis of the
given objective and associated constraint functions, the next step is to solve
Sikkim Manipal University
Page No. 39
Operations Research
Unit 3
the problem and obtain the best possible or the optimal solution using
various mathematical and analytical techniques that can be employed for
solving the linear programming model.
The graphical solution procedure is one of the methods of solving two
variable linear programming problems.
Objectives:
After studying this unit, you should be able to:
solve linear programming problems graphically
identify some exceptional cases involved in LPP
analyse the important geometric properties of LPP
Page No. 40
Operations Research
Unit 3
Any point on the line 4x1+3x2 = 210 or inside the shaded portion will satisfy
the restriction of the inequality, 4x1+3x2 <= 210. Similarly, the line
2x1+x2 = 90 meets the x1-axis at the point C(45, 0) and the x2 axis at the
point D(0, 90) as shown in figure 3.2.
Page No. 41
Operations Research
Unit 3
Figure 3.3 depicts how the area can be sketched by combining the two
graphs.
2.
Terms
Definitions
Feasible
Region
Convex
Page No. 42
Operations Research
Unit 3
3.
Half Plane
4.
Convex
Polygon
A convex polygon is a convex set formed by the intersection of finite number of closed half-planes. Refer figure
3.4 and figure 3.5
5.
Redundant
Constraint
6.
Basic
Solution
7.
Basic
Feasible
Solution
8.
Optimal
Feasible
Solution
Page No. 43
Operations Research
Unit 3
Page No. 44
Operations Research
Unit 3
Source: http://www.mirrorservice.org/sites/home.ubalt.edu/ntsbarsh/
Business-stat/opre/partVIII.htm
Fig. 3.6: Classification of Feasible Region Points
Page No. 45
Operations Research
Unit 3
Page No. 46
Operations Research
Unit 3
Co-ordinates (x, y)
(15, 0)
(40,0)
(4,18)
(6,12)
The minimum value of the objective function Z = 240 occurs at the extreme
point D (6, 12). Hence the optimal solution is x = 6, y = 12, and min Z = 240.
Page No. 47
Operations Research
Unit 3
A unique optimal solution - In this case, only one optimum solution will
be obtained in a graphical solution method.
Multiple optimal solutions/Alternative solution -: When the objective
function is parallel to a binding constraint (i.e., a constraint that is
satisfied as an equation by the optimal solution), the objective function
will assume the same optimum value at more than one solution point.
For this reason, they are called alternative optima.
An unbounded solution - When the values of the decision variables
may be increased indefinitely without violating any of the constraints, the
Page No. 48
Operations Research
Unit 3
Co ordinates
(x, y)
O
A
B
(0, 0)
(4,0)
(3,1)
0
8
10 (Max)
(0,2.5)
10 (Max)
Since any point on the line segment BC gives the maximum value (z=10)
of the objective function, there exists an alternative optima.
Sikkim Manipal University
Page No. 49
Operations Research
Unit 3
Page No. 50
Operations Research
Unit 3
It is clear from this figure that the line representing the objective
function can be moved far even parallel to itself in the direction of
increasing z, and still has some points in the region of feasible
solutions.
Hence z can be made arbitrarily large, and the problem has no finite
maximum value of z. Such problems are said to have unbounded
solutions.
Infinite profit in practical problems of linear programming cannot be
expected. If LPP has been formulated by committing some mistake, it
may lead to an unbounded solution.
3.5.3 Solved problem for inconsistent solution
Solved problem
Solve the given LPP using graphical method
Maximise z= 3x1-2x2
Subject to constraints x1+x2 1, 2x1+2x2 4, and x1,x20
Solution: Figure 3.11 depicts the problem graphically.
The figure shows that there is no point in (x1, x2) which satisfies both the
constraints simultaneously. Hence the problem has no solution because
the constraints are inconsistent.
Sikkim Manipal University
Page No. 51
Operations Research
Unit 3
This fact can be expressed more precisely by saying that the line joining
any two points in the region also lies in the region.
There are corners or extreme points on the boundary and there are
edges joining various corners.
If the optimal solution is not unique, there are points other than corners
that are optimal but in any case at least one corner is optimal.
Infeasible region.
Page No. 52
Operations Research
Unit 3
3.7 Summary
Let us recapitulate the important concepts discussed in this unit:
Any LPP with two decision variables can be solved graphically.
Any non-negative solution satisfying all the constraints is known as a
feasible solution of the problem.
The collection of all feasible solutions is known as a feasible region.
The feasible region of an LPP is a convex set.
The value of the decision variables, which maximises or minimises the
objectives function, is located on the extreme point of the convex set
formed by the feasible solutions.
Sometimes the problem may be infeasible indicating that no solution
exists for the problem.
3.8 Glossary
Inequality constraints: condition of LPP that does not hold exactly
Constraint functions: function that describes the set of conditions of LPP
Optimal solution: best solution from all feasible solutions
Page No. 53
Operations Research
Unit 3
3.10 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
Feasible
Half-plane
True
False
False
True
True
True
Terminal Questions
1. x1 = 20/19, x2 = 45/19, and max z = 235 / 19. Refer 3.4
2. Any point on the line segment BC gives the maximum value (z=10) of
the objective function. There exists an alternative solution. Refer 3.4
3. Unbounded solution. Refer 3.5.2
Page No. 54
Operations Research
Unit 3
the entire trip. The price per person and the cost for the three packages are
as follows:
Table 3.4: Summary of the Package Prices and Expenses of the Travel Agent
Tour plan
Price(Rs.)
Hotel cost(Rs.)
Meals and
expenses(Rs.)
Deluxe
10,000
4,750
3,000
Standard
7,000
2,500
2,200
Economy
6,500
2,200
1,900
In planning the trip, the following considerations must be taken into account:
At least 10% of the packages must be of the deluxe type.
At least 30% but not more than 70% must be of the standard type.
At least 30% must be of the economy type.
The maximum number of deluxe packages available in any aircraft is
restricted to 60.
The hotel desires that at least 120 tourists should be for the deluxe and
standard packages together.
The travel agent wishes to determine the number of packages to offer in
each type of trip so as to maximise the total profit.
Discussion Questions:
1. Formulate the problem as LPP.
[Hint: x1= no. of deluxe packages, x2= no. of standard packages & x3=
no. of economy packages].
2. Find the optimum solution using graphical methods for the LPP and
interpret your results.
[Hint: Plot constraint lines and mark feasibility region]
References:
Kanti Swarup & Gupta P. K., & Hira D. S., & Manmohan (2004). Operation
Research. Sultan Chand and Sons.
Page No. 55
Operations Research
Unit 3
E- References:
Linear Optimization http://www.mirrorservice.org/sites/
home.ubalt.edu/ntsbarsh/Business-stat/opre/partVIII.htm.
Page No. 56
Operations Research
Unit 4
Unit 4
Simplex Method
Structure:
4.1 Introduction
Objectives
4.2 Standard Form of LPP
Fundamental theorem of LPP
4.3 Solution of LPP Simplex Method
Initial basic feasible solution of an LPP
To solve an LPP in canonical form by simplex method
4.4 The Simplex Algorithm
Steps
4.5 Penalty Cost Method or Big M-method
4.6 Two Phase Method
4.7 Solved Problems on Minimisation
4.8 Summary
4.9 Glossary
4.10 Terminal Questions
4.11 Answers
4.12 Case Study
4.1 Introduction
In the previous unit we dealt with graphical analysis of linear programming
problems. We discussed some basic definitions, graphical methods to solve
LPP, some exceptional cases, advantages, limitations, and important
geometric properties of LPP.
In this unit, we will deal with the simplex method, which focuses on solving
an LPP of any enormity involving two or more decision variables.
The simplex algorithm is an iterative procedure for finding the optimal
solution to a linear programming problem. The objective function controls
the development and evaluation of each feasible solution to the problem. If a
feasible solution exists, it is located at a corner point of the feasible region
determined by the constraints of the system.
The simplex method simply selects the optimal solution amongst the set of
feasible solutions of the problem. The efficiency of this algorithm is because
it considers only those feasible solutions which are provided by the corner
points, and that too not all of them.
Sikkim Manipal University
Page No. 57
Operations Research
Unit 4
Objectives:
After studying this unit, you should be able to:
create a standard form of LPP from the given problem
apply the simplex algorithm to the system of equations
find solution using big M-technique
explain the importance of the two phase method
Maximise or Minimise z C i x i
i 1
Subject to:
n
a ij x j S i b i ( b i 0 )i 1, 2.......... .m.
j 1
and xj 0, j = 1, 2, --- n
S i 0, i = 1, 2, --- m
Sikkim Manipal University
Page No. 58
Operations Research
Unit 4
Page No. 59
Operations Research
Unit 4
Z = 4x1 + x2
Subject to
3x1 + x2 = 3
4x1 + 3x2 6
x1 + 2x2 3
x1, x2 0
Page No. 60
Operations Research
Unit 4
Z = (4 1 0 0 M M)
and
3 1 0 0 1 0
4 3 1 0 0 1
1 2 0 1 0 0
Subject to
x1
x2
S1
S2
A1
A
2
x1
x2
S1
S2
A1
A
2
x1
x2
S1
S2
A1
A
2
3
6
3
Page No. 61
Operations Research
Unit 4
Page No. 62
Operations Research
Unit 4
Case (iii)
Let x1 be non-basic i.e., x1= 0
1
3
The solution non-degenerates, but is not feasible.
Solution is x2 = 5/3 , x3 =
2.
3.
4.
Slack and artificial variables will form basic variable for the first
simplex table. Surplus variable will never become basic variable for
the first simplex table.
5.
Select the most negative value of Zj - Cj. That column is called key
column. The variable corresponding to the column will become basic
variable for the next table.
6.
7.
The element that lies both on key column and key row is called
pivotal element.
8.
9.
10.
11.
Page No. 63
Operations Research
Unit 4
Page No. 64
Operations Research
Unit 4
Page No. 65
Operations Research
Unit 4
Case-let
A manufacturing company discontinued production of an unprofitable
product line. This created excess production capacity. The companys
management is considering devoting this excess capacity to one or more
of the other ongoing products.
Knowing the capacity of the machines, the number of machine hours
required to produce one unit of each product and the unit profit per
product, the management needs to find out how much of each product
the company should produce to maximise profit.
The management can use the simplex method to arrive at the required
solution.
Solved problem 3
Maximise z = x1+ 9x2 + x3
Subject to x1 + 2x2 + 3x3 9
3x1 + 2x2 + 2x3 15
x1, x2, x3 0
Rewriting in the standard form
Maximise z = x1 + 9x2 + x3 + 0.S1 + 0.S2
Subject to the conditions
x1 + 2x2 + 3x3 + S1 = 9
3x1 + 2x2 + 2x3 + S2 = 15
x1, x2, x3, S1, S2 0
Where S1 and S2 are the slack variables.
Solution: The initial basic solution is x1, x2, x3 =0, S1 = 9, S2 = 15
S1
Therefore, X0 =
, C0 =
S
2
0
0
Page No. 66
Operations Research
Unit 4
X2
9
X3
1
S1
0
S2
0
S1 0
2*
S2 0
15
Zj C j
Ratio
9
= 4.5
2
15
= 7.5
2
X2
9
X3
1
S1
0
S2
0
Ratio
X2 9
1
2
3
2
1
2
9
2
S2 0
Zj C j
7/2
25
2
9
2
81
2
Since all the elements of the last row are non-negative, the optimal
81
solution is obtained. The maximum value of the objective function Z is ,
2
9
achieved for x2=
S2 = 6, are the basic variables (refer the shaded
2
column of table 4.2.). All other variables are non-basic.
Sikkim Manipal University
Page No. 67
Operations Research
Unit 4
Solved problem 4
Use simplex method to solve the LPP
Maximise
Z = 2x1 + 4x2 + x3 + x4
Subject to
x1 + 3x2 + x4 4
2x1 + x2 3
x2 + 4x3 + x4 3
x1, x2, x3, x4 0
Rewriting in the standard form
Maximise
Z = 2x1 + 4x2 + x3 + x4 + 0.S1 + 0.S2 + 0.S3
Subject to
x1 + 3x2 + x4 + S1 = 4
2x1 + x2 + S2 = 3
x2 + 4x3 + x4 + S3 = 3
x1, x2, x3, x4,
S1, S2, S3 0
Solution:
The initial basic solution is x1, x2, x3, x4 = 0, S1 = 4, S2 = 3, S3 = 3
S1
4
Therefore, X0 = S 2 = 3 C0 =
S
3
3
0
0
0
X2
X3
X4
S1
S2
S3
Ratio
S1 0
3*
S2 0
S3 0
Zj C j
3
3
1
3
1
pivot column
Sikkim Manipal University
* pivot element
Page No. 68
Operations Research
Unit 4
S1 is the outgoing variable and x2 is the incoming variable to the basic set.
The first iteration gives the table 4.4:
Table 4.4: The First Iteration Table - X2 is the Incoming Variable
X1
2
X2
4
X3
1
X2 4
1
3
S2 0
5
3
4*
S3 0
Zj C j
1
3
2
X4
1
S1
0
1
3
1
2
3
1
3
1
3
1
1
3
4
3
S2
0
S3
0
Ratio
4
3
5
3
5
3
16
3
12
pivot column
X3 enters the new basic set replacing S3, the second iteration gives the
table 4.5:
Table 4.5: Second Iteration Table X3 Replaces S3
X1
X2
X3
X4
S1
S2
S3
X2 4
1
3
1
3
1
3
4
3
S2 0
5
*
3
1
3
x3 1
1
12
1
6
Zj Cj
3
4
1
2
Ratio
1
3
5
3
1
12
1
4
5
12
5
4
1
4
23
4
pivot column
X1 enters the new basic set replacing S2, the third iteration gives the
table 4.6:
Page No. 69
Operations Research
Unit 4
X2
4
X3
1
X4
1
S1
0
X2 4
2
5
2
5
X1 2
X3 1
3
20
Zj Cj
27/20
1
5
S2
0
S3
0
Ratio
1
5
1
5
1
10
3
5
1
20
1
4
1
2
11
10
9
20
1
4
13
2
Since all the elements of the last row are non-negative, the optimal
13
1
solution is Z =
, which is achieved for x2 = 1, x1 = 1, x3 = , and x4 = 0.
2
2
Case Study
A manufacturing firm has discontinued production of a certain
unprofitable product line. This created considerable excess of production
capacity. The management is considering devoting this excess capacity
to one or more of the three products, called product 1, 2, and 3. Table 4.7
depicts the available capacity on the machines, which might limit output.
Table 4.7: Capacity of the Machine for Production
Machine Type
Available Time
(in machine hours per week)
Milling Machine
250
Lathe
150
Grinder
50
Table 4.8 depicts the number of machine-hours required for each unit of
the respective product.
Page No. 70
Operations Research
Unit 4
Product 1
Product 2
Product 3
Milling Machine
Lathe
Grinder
The unit profit would be Rs.20, Rs.6, and Rs.8 for products 1, 2, and 3.
Find how much of each product the firm should produce in order to
maximise profit?
Let x1, x2, and x3 units of products 1, 2, and 3 be produced in a week.
Then the total profit from these units is
Z = 20 x1 + 6 x2 + 8 x3
To produce these units, the management requires
8x1 + 2x2 + 3x3 machine hours of milling machine
4x1 + 3x2 + 0 x3 machine hours of lathe
and
Since the time available for these three machines are 250, 150, and 50
hours respectively, we have
8x1 +2x2 + 3x3 250
4x1 + 3x2 150
2x1 + x3 50
Obviously
x1, x2, x3 0
Subject to
Page No. 71
Operations Research
Unit 4
250
150
50
X2
X3
S1
S2
S3
Ratio
20
S1 0
250
250/8 = 31.25
S2 0
150
150
= 37.5
4
S3 0
2*
50
50
= 25
2
ZJ Cj
20
Page No. 72
Operations Research
Unit 4
X1 enters the basic set of variables replacing the variable S3. The first
iteration gives the table 4.10:
Table 4.10: The First Iteration Table- X1 Replaces S3
X1
X2
X3
S1
S2
S3
Ratio
20
S1 0
50
50
= 25
2
S2 0
3*
50
50
X1 20
1
2
1
2
25
Zj Cj
10
500
X2
X3
S1
S2
S3
Ratio
20
S1 0
1
3
2
3
8
3
50
3
50
X2 6
2
3
1
3
2
3
50
3
X1 20
1
*
2
1
2
25
50
Zj Cj
600
Page No. 73
Operations Research
Unit 4
2
3
4
3
X2
6
X3
8
S1
0
S2
0
2
3
1
3
S3
0
Ratio
50
X3 8
50
Zj c j
700
Page No. 74
Operations Research
Unit 4
table will contain the penalty cost M. You can make the following
modifications in the simplex method to minimise the error of incorporating
the penalty cost in the objective function. This method is called big Mmethod or penalty cost method.
Page No. 75
Operations Research
Unit 4
5) If non-zero artificial variables are present in the final basic set, then the
program has no solution. In contrast, zero valued artificial variables in
the final solution may exist when one or more of the original constraint
equations are redundant.
Solved problem 5
Use penalty cost method to
Maximise
z = 2x1 + 3x2
Subject to
x1 + 2x2 2
6x1 + 4x2 24
x1, x2 0
Solution: Rewriting in the standard form, we have
Maximise
Subject to
x1 + 2x2 + S1 = 2
6x1 + 4x2 S2 + A1 = 24
x1, x2, S1, S2, A1 0
X2
S1
S2
A1
S1 0
1*
2
2
1
A1 M
24
24
4
6
6M
4M
Ratio
24M
Work column
Page No. 76
Operations Research
Unit 4
X2
S1
S2
A1
X1 2
Ai M
-6
12
8M
6M
1M
12M
The process is complete as all the elements of the last two rows are nonnegative. But existence of non-zero artificial variables in the basic
solution shows that the problem has no solution (infeasible).
Self Assessment Questions
9. The value of artificial variable is M. (True/False)
10. Artificial variables are entered as basic variables. (True/False)
Page No. 77
Operations Research
Unit 4
Solved problem 6
Use the two phase method to
Maximise
z = 3x1 x2
Subject to
x1 + 3x2 2
x2 4,
x1, x2 0
2x1 + x2 2
Subject to
2x1 + x2 S1 + A1 = 2
x1 + 3x2 + S2 = 2
x2 + S3 = 4,
x1, x2, S1, S2, S3, A1 0
Phase 1: Consider the new objective,
Maximise
Z* = A1
Subject to 2x1 + x2 S1 + A1 = 2
x1 + 3x2 + S2 = 2
x2 + S3 = 4,
x1, x2, S1, S2, S3, A1 0
Table 4.15 depicts the initial simplex table arrived through the simplex
method.
Table 4.15:The Initial Simplex Table
X1
0
X2
0
S1
0
A1
1
S2
0
S3
0
A1 1
2*
S2 0
S3 0
0
2
1
1
0
1
0
0
0
0
1
0
4
2
Work column
Ratio
2
=1
2
2
=2
1
* pivot element
Page No. 78
Operations Research
Unit 4
X2
0
S1
0
A1
1
1
2
X1 0
S2 0
5
2
S3 0
0
0
1
0
0
0
S2
0
S3
0
(1/2)
0
1
0
0
1
0
4
0
Phase I is complete since there are no negative elements in the last row.
The optimal solution of the new objective is Z* = 0.
Phase II:
Consider the original objective function,
Maximise
z = 3x1 x2 + 0S1 + 0S2 + 0S3
S
x
Subject to
x1 + 2 1 =1
2
2
S1
5
x2 +
+ S2 =1
2
2
x2 + S3 = 4
x1, x2, S1, S2, S3 0
Table 4.17 depicts the corresponding simplex table with the initial
solution x1 = 1, S2 = 1, S3 = 4.
Table 4.17: The Corresponding Simplex Table
X1
3
X1 3
X2
1
1
2
S1
0
S2
0
S3
0
1
2
Ratio
1
1
2 *
S2 0
5
2
S3 0
5
2
-3/2
1
= 2
1
2
Page No. 79
Operations Research
Unit 4
Work column
* pivot element
X2
S1
S2
S3
X1 3
S1 0
S3 0
10
Since all the elements of the last row are non-negative, the current
solution is optimal.
The maximum value of the objective function is Z = 6, which is attained
for X1 = 2, X2 = 0.
Solved problem 7
Maximise
z = 3x1 + 2x2,
Subject to
2x1 + x2 2,
3x1 + 4x2 S2 + A1 12,
x1, x2 0
Rewriting in the standard form,
Maximise
Subject to
2x1 + x2 + S1 = 2
3x1 + 4x2 S2 + A1 = 2
x1, x2, S1, S2, A1 0
Solve the above using the two phase method.
Phase I:
Consider the new objective function
Maximise
z* = A1
Subject to
2x1 + x2 + S1 =2
Page No. 80
Operations Research
Unit 4
X2
0
S1
0
S2
0
A1
1
Ratio
S1 0
1*
A1 1
12
12
2
=2
1
12
=3
4
Work column
The first iteration gives the following table:
Table 4.20: The First Iteration Table
X2 0
A1 1
X1
0
2
5
5
X2
0
1
0
0
S1
0
1
4
4
S2
0
0
1
1
A1
1
0
1
0
2
4
4
Since all the elements of the last row are non-negative, the procedure is
complete. But the existence of non-zero artificial variables in the basic
set indicates that the problem is infeasible
Page No. 81
Operations Research
Unit 4
-3
-8
-M
-M
C.B
B.V
X1
X2
S1
S2
A1
A3
Qty
Ratio
-M
A1
200
200
-M
A2
1
P.E
-1
80
80 K R
S2
60
-2M +3 K C
-M + 8
Zj - C j
-3
-8
-M
C.B
B.V
X1
X2
S1
S2
A1
Qty
Ratio
Transformation
-M
A1
120
120
R1 = R1-R2
-3
x1
-1
80
R2 = R2
S2
1
P.E
60
60 K R
R3 = R3
Zj - C j
-M + 8 K C -M + 3
Cj
B.V
-3
X1
-8
X2
-M
A1
-3
-8
x1
x2
1
0
0
1
Zj - C j
0
S1
1
P.E
-1
0
0
S2
A1
Qty
Ratio
Transformation
-1
60
60
R1 = R1
0
1
0
0
80
60
-ve
R2 = R2
1
1
R3 = R3 R1
-M+3
KC
M-8
1
1
Page No. 82
Operations Research
Unit 4
-3
-8
C.B
B.V
X1
X2
S1
S2
Qty
Ratio
Transfor-mation
S1
-1
60
- ve
R1 = R1
-3
x1
-1
140
-ve
R2 = R2 + R1
-8
x2
1
P.E
60
80 K R
R3 = R1
-5
Zj - C j
-3
-8
C.B
B.V
X1
X2
S1
S2
Qty
S1
120
R1 = R1 + R3
-3
x1
200
R2 = R2 + R3
S2
Zj - C j
60
Ratio
Transformation
1
1
1
1
R3 = R3
4.8 Summary
Let us recapitulate the important concepts discussed in this unit:
The simplex algorithm is an iterative procedure for finding the optimal
solution to a linear programming problem
The simplex method simply selects the optimal solution amongst the set
of feasible solutions of the problem.
The inequality constraints of equations are obtained by adding or
subtracting the left-hand side of each such constraint by a non-negative
variable
The introduction of slack and surplus variables does not alter the
equations of constraints or the objective function.
4.9 Glossary
Slack variable: variable added to the left of type of constraint to change
it to equality.
Sikkim Manipal University
Page No. 83
Operations Research
Unit 4
Surplus variable: variable subtracted from the left of the right type of
constraint to change it to equality.
Simplex Table: table used in the simplex algorithm to keep trace of
computations made at every iteration.
4.11 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
False
True
True
True
True
True
False
True
True
True
Page No. 84
Operations Research
Unit 4
Terminal Questions
1. Rewrite in the standard form. The maximum value of the objective
function is Z = 9 which is attained for X1 = 3, X2 = 0. Refer 4.6
2. In the standard form minimize Z, Optimum solution is Z = 5.8, X1= 8/3,
X2 = 6. Refer 4.7
Page No. 85
Operations Research
Unit 4
Peak Output
(million of
kilowatt)
Investment
Cost
(Rs. 000)
Discounted
Total cost
(Rs. 000)
Steam
0.15
0.20
1200
2600
Hydroelectric:
no reservoir
0.10
0.10
1600
1680
Hydroelectric:
small reservoir
0.10
0.40
2400
2560
Hydroelectric :
large reservoir
0.80
0.90
4000
4400
Type
Discussion Questions:
1. Help the company in developing a 10-year plan that would detail the
capacity of each type of plant that it should build.
2. Develop an LP model and solve it to minimise the total discounted cost.
However, there is a restriction that no more than Rs.14,000 million can
be used for investment in plants over the 10 years.
References:
E- Reference :
Page No. 86
Operations Research
Unit 5
Unit 5
Structure:
5.1 Introduction
Objectives
5.2 Importance of Duality Concepts
5.3 Formulation of Dual Problem
5.4 Economic Interpretation of Duality
Economic interpretation of dual variables
5.5 Sensitivity Analysis
Changes in Cj of a non-basic variable
Change in Cj of a basic variable
Change in available resources
Calculating the range
5.6 Summary
5.7 Glossary
5.8 Terminal Questions
5.9 Answers
5.10 Case Study
5.1 Introduction
In the previous unit, simplex method, we studied about the standard form
and the solution for linear programming problem. The simplex method
focuses on solving linear programming problem of any enormity involving
two or more decision variables. The simplex algorithm is an iterative
procedure for finding the optimal solution to a linear programming problem.
The objective function controls the development and evaluation of each
feasible solution to the problem. If a feasible solution exists, it is located at a
corner point of the feasible region determined by the constraints of the
system.
In this unit, duality in linear programming problem, we will study about the
importance of duality concepts and the formulation of dual concepts. We will
also learn about the sensitivity analysis. Every Linear Programming Problem
(LPP) is associated with another linear programming problem involving the
same data and optimal solutions. The two problems are said to be duals of
each other. One problem is called the primal, while the other problem is
called the dual.
Sikkim Manipal University
Page No. 87
Operations Research
Unit 5
The dual formulation is derived from the same data and solved in a manner
similar to the original 'primal' formulation. In other words, you can say that
dual is the 'inverse' of the primal formulation due to the following reasons:
If the primal objective function is 'maximisation' function, then the dual
objective function is 'minimisation' function and vice-versa.
The column coefficient in the primal constraint is the row coefficient in
the dual constraint.
The coefficients in the primal objective function are the Right Hand Side
(RHS) constraint in the dual constraint.
The RHS column of constants of the primal constraints becomes the
coefficient of the dual objective function.
The concept of duality is useful to obtain additional information about the
variation in the optimal solution. These changes could be affected in the
constraint coefficient, in resource availabilities and/or objective function
coefficient. This effect is termed as post optimality or sensitivity analysis.
Objectives:
After studying this unit, you should be able to:
Page No. 88
Operations Research
Unit 5
When the primal problem has a degenerate optimal solution, the dual
has multiple optimal solutions.
When the primal problem has multiple optimal solutions, the optimal dual
solution is degenerate.
When the primal problem is unbounded, the dual is infeasible.
When the primal problem is infeasible, the dual is unbounded or
infeasible.
Page No. 89
Operations Research
Unit 5
Page No. 90
Operations Research
Unit 5
Solved problem 3
Write the dual of
Max Z = 100x1 + 200x2
Subject to 3x1 - 10 x2 15
+4x1 +15 x2 20
x1, x2, 0
Solution
First we convert
3x1 - 10x2 15
to type as shown here.
-3x1 + 10x2 -15
Therefore, its dual is:
Min Z = -15y1 + 20 y2
Subject to -3y1 + 4y2 100
10y1 + 15y2 200
y1, y2, 0
Solved problem 4
When the constraints contain = sign, write the dual of
Max Z = 40x1 + 30x2
Subject to 10x1 +6 x2 15
5x1 +7x2 -10
x1 + x2 = 9
x1 + x2 10
x1, x2, 0
Solution
Firstly,
5x1 7x2 -10
Sikkim Manipal University
Page No. 91
Operations Research
Unit 5
is rewritten as
-5x1 +7x2 10
Secondly,
x1 + x2 = 9
is written as
x1 + x2 9
Also
x1 + x2 9 this is same as -x1 - x2 -9
Therefore, the given problem is:
Max Z = 40x1 + 30x2
Subject to 10x1 + 6 x2 15
- 5x1 +7 x2 10
x1 +x2 9
- x1 x2<=- 9
-x1 x2 -10
Therefore, its dual is:
Min W = 15y1 + 10 y2 + 9y31 9y311 -10y4
Subject to 10y1 - 5y2 + y31 y311 y4 40
6y1 + 7y2 + y31 y311-y4 30
y1, y2, y31 y311,y4 0
Let y31 y311= y3
Then the dual becomes
Min W =- 15y1 +10 y2 +9 y3 -10y4
Subject to 10y1 - 5y2 + y3-y4 40
6y1 + 7y2 + y3 +y4 30
y1, y2,y4 0
y3 is unrestricted in sign
Page No. 92
Operations Research
Unit 5
Solved problem 5
Write the dual of
Min Z = 12x1 + 15x2
Subject to 5x1 +3 x2 10
x1 + x2 5
x1 0
x2 is unrestricted in sign
Solution
Let x2 =x21 x211 0. Therefore, its standard form is:
Min Z = 12x1 + 15x21 -15x211
Subject to 5x1 +3(x21- x211) 10
- [x1 +(x21- x211)] -5
Or - x1 - x21+ x211 -5
x1, x21, x211 0
Therefore, dual is:
5y1 y2 12 .(1)
3y1 y2 15 .(2)
-3y1 + y2 - 15 (3)
No.3 constraint is 3y1 y2 15 (4)
Constraints 2 and 4 give
3y1 y2 = 15
The dual problem is:
Max W = 10y1 5 y2
Subject to 5y1 y2 12
3y1 y2 =15
y1 0 y2 unrestricted in sign
Solved problem 6
Write the dual of the following linear programming problem:
Minimise Z = 3x1 2x2 + 4x3
Subject to 3x1 + 5x2 + 4x3 7
6x1 + x2 + 3x3 4
7x1 - 2x2 - x3 10
x1 - 2x2 + 5x3 3
4x1 + 7x2 - 2x3 2,
x1, x2, x3 0
Sikkim Manipal University
Page No. 93
Operations Research
Unit 5
Solution
Since the problem is of minimisation, all constraints should be of type.
Therefore, you have to multiply the third constraint throughout by -1 so
that -7x1 + 2x2 + x3 -10.
Let y1, y2, y3, y4 and y5 be the dual variables associated with the above
five constraints. Then the dual problem is given by:
Maximise W = 7y1 + 4 y2 10 y3 + 3 y4 + 2y5
Subject to 3y1 + 6y2 7 y3 + y4 + 4y5 3
5y1 + y2 + 2y3 2y4 + 7y5 -2
4y1 + 3y2 + y3 + 5y4 2y5 4
y1, y2, y3, y4, y5 0
Solved problem 7
Find the dual of:
Maximise 12x1 + 10x2
Subject to 2x1 + 3x2 18
2x1 + x2 14
x1, x2 0
Solution
The 'dual' formulation for this problem will be as follows:
Minimise 18y1 + 14y2
Subject to 2y1 + 2y2 12
3y1 + y2 10
y1>= 0, y2 0
Points to be noted
The column coefficients in the primal constraint namely (2, 2) and
(3, 1) have become the row coefficients in the dual constraints.
The coefficient of the primal objective function namely 12 and 10
have become the constants in the right hand side of the dual
constraints.
The constants of the primal constraints, namely 18 and 14, have
become the coefficient in the dual objective function.
The direction of the inequalities has been reserved.
While the primal is a maximisation problem, the dual is a
minimisation problem.
Sikkim Manipal University
Page No. 94
Operations Research
Unit 5
Solved problem 8
Obtain the dual problem of the following primal formulation:
Maximise Z = 2x1 + 5x2 + 6x3
Subject to 5x1 + 6x2 -x3 3
-2x1 + x2 + 4x3 4
x1 - 5x2 + 3x3 1
-3x1 - 3x2 + 7x3 6
x1, x2, x3 0
Solution
Step 1 - Write the objective function of the dual. As there are four
constraints in the primal, the objective function of the dual will have
4 variables.
Minimise r* = 3 w1 + 4 w2 + w3 + 6 w4
Step 2 - Write the constraints of the dual. As all the constraints in the
primal are '<', the constraints in the dual will be '>'. The column COM+
efficient of the primal becomes the row coefficient of the dual.
Constraints 5w1 - 2w2 + w3 - 3w4 2
6w1 + w2 - 5w3 -3w4 5
-w1 + 4w2 + 3w3 + 7w4 6
Step 3 - Therefore the dual of the primal is:
Minimise Z = 3w1 + 4w2 + w3 + 6w4
Subject to: 5w1 - 2w2 + w3 - 3w4 2
6w1+ w2 - 5w3 - 3w4 5
- w1 + 4w2 + 3w3 + 7w4 6
w1, w2, w3, w4 >= 0
Solved problem 9
Obtain the dual of the following linear programming problem:
Minimise Z = 5x1 -6x2 + 4x3
Subject to the constraints: 3x1 + 4x2 + 6x3 9
x1 + 3x2 + 2x3 5
7x1 - 2x2 - x3 10
x1 - 2x2 + 4x3 4
2x1 + 5x2 - 3x3 3
x1, x2, x3
0
Sikkim Manipal University
Page No. 95
Operations Research
Unit 5
Solution
As you can see, one of the primal constraints is a " constraint while the
others are all " constraints. The dual cannot be worked out unless all
the constraints are in the same direction. To convert this into "
constraint, multiply both the sides of the equation by "-" sign. After
multiplying the constraint by "-" sign, it will become 7x1 + 2x2 + x3 -10.
Now that all the constraints are in the same direction and the dual can be
worked out, the dual formulation is:
Maximise Z = 9w1 + 5w2 - 10w3 + 4w4 + 3w5
Subject to: 3w1 + w2 - 7w3 + w4 + 2w5 5
4w1 + 3w2 + 2w3 - 2w4 + 5w5 -6
6w1 + 2w2 + w3 + 4w4 + -3w5 4
w1, w2, w3, w4, w5
0
Self Assessment Questions
3. The coefficients of decision variables in the objective function become
quantities on the right hand side of ___________________.
4. constraints changes to ________ type in dual linear programming.
5. For every linear programming problem, there exists a unique
________ problem.
Page No. 96
Operations Research
Unit 5
Dual
Maximise
Minimise
c j. x j
bi . y i
i1
j1
Subject to
Subject to
n
aij x j bi , i 1, 2, ..., m.
aij y i c i , j 1, 2, ..., n.
xj 0, j = 1,2,., n
yj 0, i = 1,2,., m
j1
i1
j1
cj xj
bi yi w
i1
The strict equality, z = w, holds true when both the primal and dual solutions
are optimal. Consider the optimal condition z = w. Given that the primal
problem represents a resource allocation model, you can think of z as
representing the profit in rupees. bi represents the number of units available
of the resource i. Therefore, we can express the equation z = w as
profit (Rs) = (units of resource i) x (profit per unit of resource i).
This means that the dual variables yi, represent the worth per unit of
resource i.
Sikkim Manipal University
Page No. 97
Operations Research
Unit 5
Note: Variables yi are also called as dual prices, shadow prices and simplex
multipliers.
With the same logic, the inequality z < w associated with any two feasible
primal and dual solutions is interpreted as (profit) < (worth of resources).
This relationship implies that as long as the total return from all the activities
is less than the worth of the resources, the corresponding primal and dual
solutions are not optimal. Optimality is achieved only when the resources
have been exploited fully, which can happen when the input equals the
output (profit). Economically, the system is said to be disequilibrium (non
optimal) when the input (worth of the resources) exceeds the output (return).
Equilibrium occurs only when the two quantities are equal.
Self Assessment Questions
6. Dual variables represent the worth or unit of a resource. (True/False)
7. Optimality is reached when the resources are not fully utilised.
(True/False)
8. At optimum level, the relationship holds as a strict equation.
(True/False)
Page No. 98
Operations Research
Unit 5
with various values of the crucial data, to get an idea of the range of
possible outcomes.
Sensitivity analysis is important due to the following reasons:
values of linear programming parameters might change
linear programming parameters have an uncertainty factor attached to
them
Any change in a linear programmings parameters affects the optimality of
the basic variable. This is due to the following two reasons:
When a variable (or number of variables) in row 0 has a negative
coefficient, we can obtain a better basic feasible solution of larger
z-value, by pivoting a non-basic variable with a negative coefficient in
row 0. If this occurs, the basic variable becomes a sub-optimal basis.
A constraint (or number of constraints) may have a negative RHS. In
such a case, at least one member of basic variable will become negative
and will no longer yield a basic feasible solution. This is an infeasible
basic.
The following six types of changes in linear programmings parameters
cause changes in the optimal solution:
Changing the objective function coefficient of a non-basic variable.
Changing the objective function coefficient of a basic variable.
Changing the right-hand side of a constraint.
Changing the column of a non-basic variable.
Adding a new variable or activity.
Adding a new constraint.
Solved problem 10
Luminous lamps produce three types of lamps A, B and C. These lamps
are processed on three machines X, Y and Z.
To find out the optimal solution in order to maximise profit and minimise
cost, first we find out whether a previously determined optimal solution
remains optimal if the contribution rate is changed. An increase in Cj of a
variable would mean that, resources from other products should be
diverted to this more profitable product.
The reverse is true for a minimisation problem.
Sikkim Manipal University
Page No. 99
Operations Research
Unit 5
Machine
M1
M2
M3
10
12
Available time
100
77
80
Solution
The linear programming model for this problem is as follows:
Maximise Z = 12 x1 + 3 x2 + x3
Subject to10 x1 + 2 x2 + x3 100
7 x1 + 3 x2 + 2x3 77
2 x1 + 4 x2 + x3 80
x1, x2 , x3 0
On solving this problem by the simplex method, we get
73
35
x1 =
, x2 = , x3 = 0
8
8
981
And the optimal value of the objective function is
.
8
5.5.1 Changes in Cj of a non-basic variable
A non basic variable can be brought on the basis only if its contribution rate
is attractive. Therefore, we need to determine the upper limit of the profit
contribution (Cj) of each non basic variable. The reverse is true for a
minimisation problem. Table 5.3 depicts the final simplex table of this
problem.
Operations Research
Unit 5
12
CB
Basic
variables B
x1
x2
x3
s1
s2
s3
Solution values
b (=xB)
12
x1
1
16
3
16
1
8
73
8
x2
5
8
35
8
s3
9
4
177
4
zj
12
zjcj
13
16
7
16
17
8
11
8
27
16
15
16
3
8
11
16
15
16
3
8
From this final simplex table, we can see that profit contribution for product
C is Re 1, which is not greater than its zj. Therefore, to bring x3 into the
27
basis, its profit contribution rate cj must exceed Rs.
to make zjcj value
16
negative or zero that is zjcj 0. Specifically,
If cj* cj>zj cj,, then a new optimal solution must be derived.
If cj* cj = zj cj, then an alternative optimal solution exists.
If cj* cj<zj cj, then the current optimal solution remains unchanged.
In case of c3 = 1 and z3 c3 =
11
,
16
11
C3* 1
16
27
11
C3*
+1 =
16
16
Operations Research
Unit 5
Z C
; y ij 0 C i Minimum
- Minimum
y ij
Z j C j
; y ij 0
y ij
8 = minimum (11, 3) = 3
Minimum 16
,
1 1
16 8
Minimum 16
3
16
=5
Hence,
5 c1* -12 3 i.e. 7 c1* 15
Thus, we can say that the optimal solution is insensitive as long as the
changed profit coefficient c1* varies between Rs. 7 and Rs. 15.
5.5.3 Change in available resources
If the available resources change, we need to investigate whether a
previous optimal solution remains feasible or not. For long term planning, it
is important to know the bounds within which each available resource (for
instance machine hours) can vary, without causing drastic changes in the
current optimal solution. To illustrate this, divide each quantity in the XB
column by the corresponding coefficient in the X4 column of the table 5.4.
Table 5.4 depicts the resources table.
Operations Research
Unit 5
X4
XB
X4
73
8
3
16
146
3
35
8
7
16
10
177
4
11
8
354
11
354
indicates the number of hours of machine M1
The least positive ratio
11
that can be decreased. The least negative ratio (10) indicates the increase
in the number of hours of machine M1.
354
11
746
11
746
11
to 110.
We can calculate the range of hours for machine M2 and M3 in the same
manner.
Self Assessment Questions
9. Sensitivity analysis is carried out on _______ simplex table.
10. Sensitivity analysis helps us to study the effect of changes in _______
_______ in objective function.
11. The results of sensitive analysis establish ___________ and
________ ___________ for input parameters value.
Operations Research
Unit 5
5.6 Summary
Let us recapitulate the important concepts discussed in this unit:
For every linear programming problem there exists a dual linear
programming problem.
The dual linear programming problem helps us to reduce the amount of
calculation involved in original linear programming problem.
The dual linear programming problem also helps us to interpret the
economic variables more effectively.
5.7 Glossary
Optimality: most desirable solution subject to constraints
Canonical form: standard form
Primal problem: original linear programming problem
Operations Research
Unit 5
5.9 Answers
Self Assessment Questions
1. False
2. True
3. Dual
4.
5. Dual
6. True
7. False
8. True
9. Final
10. Resource levels
11. Upper, lower, bounce
Terminal Questions
1. Minimise W = 6y1 + 12y2
Subject to constraints
y1 + 4y2 7
2y1 + 3y2 5
y1 + y2 0
2. Minimise w = 200y1 + 150y2 100y3 80y4
Subject to constraints
5y1 + 3y2 5y3 8y4 3
4y1 + 5y2 4y3 4y4 4
y1, y2, y3, y4 0
3. Minimise w = 12y1 + 8y2 + 8y3
Subject to constraints
4y1 + 4y2 + 4y3 2
3y1 + y2 - y3 1
y1, y2, y3, 0
Operations Research
Unit 5
hardships due to droughts, floods, etc., are an important part of the family
history. However, the Oxmans enjoy their self-reliant lifestyle, and gain
considerable satisfaction from continuing the family tradition of successfully
living off the land during an era, when many family farms are being
abandoned or taken over by large agricultural corporations.
John Oxman is the current manager of the farm while his wife Eunice runs
the house and manages the farms finances. Johns father, grandpa Oxman,
lives with them and still puts in many hours working on the farm. John and
Eunices older children, Frank, Phyllis, and Carl, also are given heavy
chores before and after school. The entire family can produce a total of
4,000 person-hours worth of labour during the winter and spring months,
and 4,500 person-hours during the summer and fall. If any of these personhours are not needed, Frank, Phyllis, and Carl will use them to work on a
neighbouring farm for $5 per hour during the winter and spring months and
$5.50 per hour during the summer and fall.
The farm supports two types of livestock: dairy cows and laying hens, as
well as three crops: soybeans, corn, and wheat. (All three are cash crops,
but the corn is also a feed crop for data per acre planted and wheat is also
is used for chicken feed.) The crops are harvested during the late summer
and fall. During the winter months, John, Eunice, and grandpa make a
decision about the mix of livestock and crops for the coming year.
Currently, the family has just completed a particularly successful harvest,
which has provided an investment fund of $20,000 that can be used to
purchase more livestock. (Other money is available for ongoing expenses,
including the next planting of crops). The family currently has 30 cows
valued at $35,000 and 2,000 hens valued at $5,000. They wish to keep all
this livestock and perhaps purchase more. Each new cow would cost
$1,500, and each new hen would cost $3.
Over a years time; the value of a herd of cows will decrease by about 10
percent, and the value of a flock of hens will decrease by about 25 percent
due to aging. Each cow will require 2 acres of land for grazing and 10
person-hours of work per month, while producing a net annual cash income
of $850 for the family. The corresponding figures for each hen are: no
significant acreage, 0.05 person-hours/month, and an annual net cash
income of $4.25. The chicken house can accommodate a maximum of
Sikkim Manipal University
Operations Research
Unit 5
5,000 hens, and the size of the barn limits the herd to a maximum of
42 cows.
For each acre planted in each of the three crops, table 5.5 depicts the
number of person-hours of work that will be required during the first and
second halves of the year, as well as a rough estimate of the crops net
value (in either income or savings in purchasing feed for the livestock).
Table 5.5: Corns Net Value
Data per acre planted
Soybeans
Corn
Wheat
1.0
0.9
0.6
1.4
1.2
0.7
Rs. 70
Rs. 60
Rs. 40
Net value
To provide much of the feed for the livestock; John wants to plant at least
1 acre of corn for each cow in the fund. Plus the value of the livestock at the
end of the coming year, plus any income from working on a neighbouring
farm, minus living expenses of $40,000 for the year coming years herd, in
addition, at least 0.05 acre of wheat for each hen in the coming years flock.
John, Eunice, and grandpa are now discussing how much acreage should
be planted in each of the crops and how many cows and hens to have for
the coming year. Their objective is to maximise the familys monetary worth
at the end of the coming year; (the sum of the net income from the livestock
for the coming year plus the net value of the crops for the coming year plus
what remains from the investment).
Discussion Questions:
1. Identify verbally the components of a linear programming model for this
problem.
2. Formulate this model. (Either an algebraic or a spreadsheet formulation
is acceptable.)
3. Obtain an optimal solution and generate the additional output provided
for performing post optimality analysis (for example, the sensitivity report
when using excel). What does the model predict regarding the familys
monetary worth at the end of the coming year?
Operations Research
Unit 5
References:
E-References:
http://www.psnacet.edu.in/courses/MBA/Linearprogramming/lecture11.pdf
http://agecon2.tamu.edu/people/faculty/mccarl-bruce/mccspr/new04.pdf
http://www.sce.carleton.ca/faculty/chinneck/po/Chapter6.pdf.
http://www-ee.eng.hawaii.edu/~jyee/491.06/lectures/chapter06.pdf
Operations Research
Unit 6
Unit 6
Transportation Problem
Structure:
6.1 Introduction
Objectives
6.2 Formulation of Transportation Problem (TP)
6.3 Transportation Algorithm (MODI Method)
6.4 The Initial Basic Feasible Solution
North-west corner rule
Matrix minimum method
Vogels approximation method
6.5 Moving Towards Optimality
Improving the solution
Modified distribution method/MODI method/U V Method
Degeneracy in transportation problem
6.6 Summary
6.7 Glossary
6.8 Terminal Questions
6.9 Answers
6.10 Case Study
6.1 Introduction
In the previous unit, duality in linear programming problem, we studied
about how every linear programming problem (LPP) is associated with
another linear programming problem involving the same data and optimal
solutions. We also learnt about formulation of dual problem, economic
interpretation of duality and sensitivity analysis. In this unit, transportation
problem, we will study about the formulation of transportation problem. We
will learn about the initial basic feasible solution and the various methods.
We will also study how to improve the solution and about the modified
distribution method. To facilitate the presentation and solution, the general
transportation problem is normally portrayed in a tabular form.
Transportation model is an important class of linear programs. For a given
supply at each source and a given demand at each destination, the model
studies the minimisation of the cost of transporting a commodity from a
number of sources to several destinations.
Sikkim Manipal University
Operations Research
Unit 6
Cities
(Destination)
Delhi
D1
Mumbai
S1
Hyderabad
D2
Kolkata
S2
Bangalore
D3
Chennai
S3
Ahmedabad
D4
Where,
Si = Quantity of petrol available at source i, (i=1, 2, 3)
Dj = Quantity of petrol required at destination j, (j = 1, 2, 3, 4)
Cij = The cost of shipping one barrel of petrol from source i to destination j
for each route
Xij =The number of barrels of petrol per route from source i to destination j
Sikkim Manipal University
Operations Research
Unit 6
D1
S1
S2
S3
Sm
Demand
D2
C11
x11
C12
x12
C21
x21
C22
C31
b1
C32
b2
a1
C2n
a2
C3n
a3
Cm
am
x3n
Cm
xm2
C1n
x2n
x32
Cm
Supply
x1n
x22
x31
xm1
Dn
xmn
bn
ai = bi
Operations Research
Unit 6
Each source has one row and each destination has one column. The
capacity of each source is shown at the end of each row, and each
destination is written under its corresponding column. These capacities and
demands are known as rim conditions. The unit cost to ship from each
source to each destination is written in the upper-right-hand corner of the
cell in the matrix.
If the total demand is equal to the total supply it is a balanced transportation
problem. When the demand is not equal to the supply it is an unbalanced
transportation problem.
It is assumed that the total supply and the total demand are equal.
m
i 1
ai
bj
(1)
j 1
i 1
j 1
Cij x ij
Subject to
n
x ij bj ; ij 1, 2,.......... ., n
x ij a i ,i 1, 2,......,m
j1
m
i1
(2)
Operations Research
Unit 6
i1
j1
a i bj
Self Assessment Questions
1. Transportation problems are a special type of ___________.
2. The number of rows and columns need not always be ___________.
3. If total demand is greater than total supply then there is _____ ____
solution.
Operations Research
Unit 6
Operations Research
Unit 6
D2
D3
D4
Supply
O1
14
O2
16
O3
Demand
10
15
35
Where Oi and Dj represent the ith origin and the jth destination respectively.
Operations Research
Unit 6
Solution
1. Start with the top most left corner
2. Allocate the maximum possible amount here.
3. Move to the right hand cell if there is still any available quantity,
otherwise move to down cell. Again perform step 2 and continue
until all the available quantity is exhausted.
Following north west corner rule, the first allocation is made in the
cell (1,1), the magnitude being x11 = min (14, 6) = 6
The second allocation is made in the 6 10 15 4 cell (1, 2) and
the magnitude of allocation is given by x12= min (14 6, 10) = 8
The third allocation is made in the cell (2, 2), the magnitude being
x22 = min (16, 10 8) = 2.
The magnitude of fourth allocation, in the cell (2, 3) is given
by x23 = min (16 2, 15) = 14.
The fifth allocation is made in the cell (3, 3), the magnitude
being x33 = min (5, 15 14) =1.
The sixth allocation in the cell (3, 4) is given by x34 = min
(5 1, 4) = 4.
Now all the rim requirements have been satisfied and hence, an initial
feasible solution to the transportation problem has been obtained.
Table 6.3 depicts the solution.
Table 6.3 Initial Feasible Solution to the Transportation Problem
D1
O1
6
6
8
D2
4
8
9
2
3
O2
4
O3
6
10
D3
D4
2
14
6
1
15
14
16
2
4
5
4
Operations Research
Unit 6
D2
D3
D4
Capacity
O1
O2
O3
10
Demand
24
Solution
The transportation table of the given transportation problem has 12 cells.
Following the matrix minima method,
The first allocation is made in the cells (3, 1), the magnitude being x31 = 4.
This satisfies the requirement at destination D1 and thus we cross the first
column from the table 6.5. The second allocation is made in the cell
(2, 4), the magnitude being x24 = min (6, 8) =6. Cross the fourth column of
the table 6.5. This yields the table 6.5 as depicted.
Sikkim Manipal University
Operations Research
Unit 6
There is again a tie for the third allocation. Arbitrarily choose the cell
(1, 2) and allocate x12 = min (6, 6) = 6. Cross the second column of the
first row. Next, choose to cross off the first row of the table 6.5. The next
allocation of magnitude x32 = 0 is made in the cell (3, 2), cross the second
column. Table 6.7 depicts the result
Table 6.7: Third Iteration
Again you choose arbitrarily to make the next allocation in cell (2, 3) of
magnitude x23 = min (2, 8) = 2, cross the second row in table 6.7. The last
allocation of magnitude x33 = min (6, 6) = 6 is made in the cell (3, 3). Now
that all the rim requirements have been satisfied, an initial feasible
solution has been determined. This solution is depicted in table 6.8.
Since the cells do not form a loop, the solution is basic and degenerate.
The transportation cost according to the above route is given by:
Z=62+22+60+40+02+62=28
6.4.3 Vogels approximation method
The Vogels approximation method (VAM) takes into account not only the
Sikkim Manipal University
Operations Research
Unit 6
least cost cij, but also the cost that just exceeds cij. The steps of the method
are given as follows:
Step 1 - For each row of the transportation table, identify the smallest and
the next to smallest costs. Determine the difference between them for each
row. Display them alongside the transportation table by enclosing them in
parenthesis against the respective rows. Similarly, compute the differences
for each column.
Step 2 - Identify the row or column with the largest difference among all the
rows and columns. If a tie occurs, use any arbitrary tie breaking choice. Let
the greatest difference correspond to the ith row and let Cij be the smallest
cost in the ith row. Allocate the maximum feasible amount xij = min (ai, bj) in
the (i, j)th cell and cross off the ith row or the jth column in the usual manner.
Step 3 - Recompute the column and row differences for the reduced
transportation table and go to step 2. Repeat the procedure until all the rim
requirements are satisfied.
Remarks
A row or column difference indicates the minimum unit penalty incurred
by failing to make an allocation to the least cost cell in that row or
column.
It is clear that VAM determines an initial basic feasible solution, which is
very close to the optimum solution, but the number of iterations required
to reach the optimal solution is small.
Area of application
It is used to compute transportation routes in such a way as to minimise
transportation cost for finding out location of warehouses
It is used to find out locations of transportation corporations depots
where insignificant total cost difference may not matter.
Operations Research
Unit 6
Solved problem 3
A company has three production facilities S1, S2 and S3 with production
capacity of 7,9 and 18 units (in 100s) per week of a product respectively.
These units are to be shipped to four warehouses D1, D2, D3 and D4
with requirement of 5, 8, 7 and 14 units (in 100s) per week, respectively.
Table 6.9 depicts the transportation costs (in rupees) per unit between
factories to warehouses.
Table 6.9: Transportation Costs
D1
D2
D3
D4
S1
19
30
50
10
S2
70
30
40
60
S3
40
70
20
Solution
The differences (penalty costs) for each row and column have been
calculated as depicted in table 6.10. In the first round, the maximum
penalty, 22 occurs in column D2. Thus the cell (S3, D2) having the least
transportation cost 8 is chosen for allocation. The maximum possible
allocation in this cell is 8 and it satisfies demand in column D2. Adjust the
supply of S3 from 18 to 10 (18-10).
The new row and column penalties are calculated except column D2
because its demand has been satisfied. The second round allocation is
made in column D1 with target penalty 21 in the same way as in the first
round as depicted in cell (S1, D1) of the table 6.10..
In the third round, the maximum penalty 50 occurs at S3. The maximum
possible allocation of 10 units is made in cell (S3, D4) having least
transportation cost 20 as depicted in table 6.10.
The process is continued with new allocations till a complete solution is
obtained. The initial solution using Vogels approximation method is
depicted in table 6.10.
The total transportation cost associated with this method is calculated as
follows:
Operations Research
Unit 6
D2
D3
D4
Supply
S1
19 (5)
30
50
10 (2)
40
40
S2
70
30
40 (7)
60 (2)
10
20
20
20
S3
40
8 (8)
70
20 (10)
18
12
20
50
Demand
14
34
21
22
10
10
21
10
10
10
10
10
50
Column
difference
row difference
Operations Research
Unit 6
Operations Research
Unit 6
I
II
III
Demand
10
15
Supply
14
12
5
Solution
Since total demand = 31 = total supply, the problem is balanced. The
initial basic feasible solution is obtained by Vogels approximation method.
Table 6.12 depicts the initial solution.
Table 6.12 Initial Solution
Supply
6
4
14
14
4
6
9
5
8
1
12
6
5
5
Demand
For allocated cells
u1 + v3 = 4
Set u2 = 0
u1 + v3 = 4
u2 + v2 = 9
10
15
u1 = -4
u3= -4
v1 = 4
Operations Research
u2 + v3 = 8
u3 + v2 = 5
Unit 6
v2 = 9
v3 = 8
Note: Variable ui/vi is repeated very often for easy calculation. Here u2 is
repeated often. For unallocated cells ij = cijuivj
11 = 6 (-4) 4 = 6
12 = 8 (-4) 9 = 3
31 = 1 (-4) 4 = 1
33 = 6 (-4) 8 = 2
The optimum allocations are x13 = 14, x21 = 6, x22 = 5, x23 = 1, x32 = 5
The minimum transportation cost is 144+64+59+18+52 = 143
Destinations
Sources
S1
S2
D1
45
14
D2
17
18
D3
21
19
D4
30
31
Operations Research
Unit 6
Solution
Since the supply and requirements are not equal it is called an
unbalanced transportation problem. In general, if aibj then it is called
an unbalanced transportation problem. We introduce either a dummy row
or a column with cost zero quantities and biaj respectively. Applying
Vogels approximation method we find the basic feasible solution.
Table 6.14 depicts the basic feasible solution.
Table 6.14: Basic Feasible Solution using Vogels Approximation Method
D1
D2
D3
D4
Supply
S1
45
17 (6)
21(3)
30 (6)
15/9/6
11
S2
14(9)
18
19 (4)
31
13/4
12
S3
0 (3)
Demand
9/0
7/3
9/6
31
14
17
19
30
31
Column
difference
=3x0
= 126
= 102
= 4 x 19
= 6 x 30
= 3 x 21
547
u1 + v4 = 30
u3 = -30
u2 + v1 = 14
v1 = 16
u2 + v3 = 19
v2 = 17
u3 + v4 = 0
v3 = 21
row difference
=
=
=
76
180
63
Set u1 = 0
u2= -2
v4 = 30
Sikkim Manipal University
Operations Research
Unit 6
Operations Research
Unit 6
6.6 Summary
Let us recapitulate the important concepts discussed in this unit:
The transportation problem is a special type of linear programming
problem in which the objective is to transport a homogeneous product
manufactured at several plants (origins) to a number of different
destinations at a minimum total cost.
There are several different techniques for computing an initial basic
feasible solution to a transportation problem, such as North-west corner
rule, Matrix minimum method and Vogels approximation method.
The degeneracy in transportation problem can develop in two ways
o The degeneracy develops while determining an initial assignment via
any one of the initial assignment methods discussed.
o The degeneracy develops at the iteration stage.
Sikkim Manipal University
Operations Research
Unit 6
6.7 Glossary
Algorithm: sequence of steps to solve complex problems
Degeneracy: In transportation problem the number of occupied cells is less
than m+n-1 where m is number of rows and n is number of columns
Factories
11
15
Operations Research
Unit 6
supply
O1
12
10
O2
O3
11
11
O4
11
10
Demand
22
6.9 Answers
Self Assessment Questions
1. Linear programming problem
2. Equal
3. No feasible
4. True
5. True
6. True
7. True
8. False
9. True
10. Zero or positive
11. Not equal
12. m + n 1
Terminal Questions
1. The optimal transportation cost is Rs. 796
2 The optimal transportation cost is Rs. 10, 000
3 The minimum transportation cost is Rs. 112 as 0
Operations Research
Unit 6
Market
Market
Source
61
72
45
55
66
31
38
24
35
69
78
60
49
56
36
43
28
24
31
59
66
63
61
47
33
36
32
26
Table 6.19 depicts the capital investment in ships required for each million
board feet to be transported annually by ship along each route.
Table 6.19: Capital Investment for Ships
Investment for Ships ($1,000's)
Market
Source
275
303
238
285
293
318
270
250
265
283
275
268
240
Considering the expected useful life of the ships and the time value of
money, the equivalent uniform annual cost of these investments is one-tenth
the amount depicted in the table 6.19. The objective is to determine the
overall shipping plan that minimises the total equivalent uniform annual cost.
Operations Research
Unit 6
You are the head of the operations research team that has been assigned
the task of determining this shipping plan for each of the following three
options:
Option 1: Continue shipping exclusively by rail
Option 2: Switch to shipping exclusively by water
Option 3: Ship by either rail or water, depending on which is less
expensive for the particular route
Discussion Questions:
1. Present your results for each option and compare.
2. Consider the fact that these results are based on current shipping and
investment costs, so the decision on the option to adopt now should take
into account, managements projection of how these costs are likely to
change in the future. For each option describe the scenario of future
cost changes that would justify adopting the option now.
References:
Kanti Swarup & Gupta P. K., & Hira D. S., & Manmohan (2004).
Operation Research. Sultan Chand and Sons.
E-References:
http://www.rajalakshmi.org/Department/MECH/2009ES/ME1018
NOL.doc. accessed on April 28, 2010.
Operations Research
Unit 7
Unit 7
Assignment Problem
Structure:
7.1 Introduction
Objectives
7.2 Mathematical Formulation of the Problem
7.3 Hungarian Method Algorithm
7.4 Routing Problem
Unbalanced AP
Infeasible assignments
Maximisation in AP
7.5 Travelling Salesman Problem
7.6 Summary
7.7 Glossary
7.8 Terminal Questions
7.9 Answers
7.10 Case Study
7.1 Introduction
In the previous unit we dealt with the formulation of Transportation Problem
(TP), transportation algorithm (MODI method), and the initial basic feasible
solution. We also discussed about how to get the optimum solution. In this
unit, we will deal with assignment problem. The assignment problem is a
special case of transportation problem, where the objective is to minimise
the cost or time of completing a number of jobs by a number of persons, and
to maximise revenue and sales efficiently.
In other words, when the problem involves the allocation of n different
facilities to n different tasks, it is often termed as an assignment problem.
This model is mostly used for planning. The assignment model is useful in
solving problems, such as assignment of machines to jobs, assignment of
salesman to sales territories, travelling salesman problem, and many more
similar situations. It may be noted that with n facilities and n jobs, there
are n possible assignments.
One way of finding an optimal assignment is to write all the n possible
arrangements, evaluate their total cost, and select the assignment model
Sikkim Manipal University
Operations Research
Unit 7
Operations Research
Unit 7
Fig. 7.1: Steps to Solve the Minimisation Case for an Assignment Problem
Operations Research
Unit 7
The result of steps 3a) and 3b) is a revised total opportunity cost matrix.
Step 4: Repeat steps 2 and 3 until an optimal assignment having a total
opportunity cost of zero can be made.
Figure 7.2 depicts the various methods used to solve the assignment
problem by any of the following four methods.
Enumeration method
Simplex method
Transportation problem
Hungarian method
However, in this unit, we will focus on the most commonly used method, that
is the Hungarian method, to solve the assignment problems.
Objectives:
After studying this unit, you should be able to:
interpret an assignment problem mathematically
explain a routing problem
describe the travelling salesman problem
state the significance of the assignment problem
compute the problem using the Hungarian method
explain practical problems, such as routing problems and travelling
salesman problems
Operations Research
Unit 7
xij = 1 and
i 1
j 1
xij = 1
j 1
i 1
xij cij
i1
j1
xij cij
xij = 1
i 1
j =1, 2, n
And
j 1
xij = 1
i = 1, 2 n
With xij = 0 or 1
Note: In an assignment problem, if you add a real number to or subtract a
real number from each element of a row or column of the cost matrix, then
the optimum assignment for the modified matrix is also optimum for the
original one.
Operations Research
Unit 7
Operations Research
Unit 7
Operations Research
Unit 7
Operations Research
Unit 7
Solved problem 1
Solve the following AP.
Table 7.4: Assignment Operations Table
Operations Research
Unit 7
Operations Research
Unit 7
Operations Research
Unit 7
Solved problem 3
Solve the following maximisation assignment problem.
Table 7.11: Assignment Table
Operations Research
Unit 7
Operations Research
Unit 7
Now apply the Hungarian method to this problem and get the row and
column reduced matrices as depicted in table 7.17 and 7.18.
Table 7.17: Row Reduction
Operations Research
Unit 7
Again draw minimum number of lines to cover all the zeros as depicted
in table 7.21.
Table 7.21: Number of Lines = Number of Rows
After the above steps, row 2 has zero as the last entry, mark it with .
Now row 1 will have the third entry zero uncovered. Mark it with . So
row 4 will have only one entry in cell (4, 4) with zero. This gives you the
optimal assignment as depicted in table 7.23.
Sikkim Manipal University
Operations Research
Unit 7
Thus, x13 = x25 = x31 = x44 = x52 = 1 and rest all xij = 0. The optimal
assignment is depicted in table 7.24.
Table 7.24: Optimal Assignment
Operations Research
Unit 7
Operations Research
Unit 7
xijk = 1
k =1, 2, 3n
i j
b) Only one other city may be reached from a specific city i, thus
xijk =1, i = 1, 2, n
c) Only one other city can initiate a direct arc to a specified city j, thus
xijk =1,
j =1, 2... n
d) Given the kth directed arc ends at some specific city j, the (k+1)th directed
arc must start at the same city j; thus
xijk =
rj
i j
These constraints ensure that the round trip will consist of connected
arcs. The objective function is to minimise
i j
z =
d
x
ij
ijk
7.6 Summary
Let us recapitulate the important concepts discussed in this unit:
This unit on assignment problems focuses on a special type of
transportation problem, where the objective is to allocate n number of
different facilities to n number of different tasks.
Sikkim Manipal University
Operations Research
Unit 7
7.7 Glossary
Opportunity cost matrix: A cost matrix that depicts the lost opportunity
when allocating resource to activity.
Job
s
J1
J2
J3
J4
M1
15
17
14
16
Machines
M2
M3
11
13
12
12
15
10
13
11
M4
15
13
14
17
Operations Research
Unit 7
3. In a plant layout, there are five vacant places. The plant orders four
machines to be installed in the vacant places. The cost of installing is as
follows:
M/G
M1
M2
M3
M4
A
9
12
14
B
11
9
11
8
C
15
14
12
D
10
10
11
7
E
11
9
7
8
4
21
15
15
12
7.9 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
True
True
False
True
True
False
Maximisation problem
Infeasible
Only once
(n-1)
Terminal Questions
1. The optimum assignment policy is
Job1 to machine 2, Job 2 to machine 4
Sikkim Manipal University
Operations Research
Unit 7
Plant
A
B
Distribution Centre
1
2
3
Rs. 800
Rs. 700
Rs. 400
Rs. 600
Rs. 800
Rs. 500
A total of 60 fork-lift trucks are produced and shipped per week. Each plant
can produce and ship any amount up to a maximum of 50 trucks per week,
so there is considerable flexibility on how to divide the total production
between the two plants so as to reduce shipping costs. However, each
distribution centre must receive exactly 20 trucks per week.
Assume that distribution centres 1, 2, and 3 must receive exactly 10, 20,
and 30 units per week respectively. For administrative convenience,
management has decided that a single plant will supply each distribution
centre totally, so that one plant will supply one distribution centre and the
other plant will supply the other two distribution centres. The choice of
assignments of plants to the distribution centres is to be made solely on the
basis of minimising total shipping cost.
Operations Research
Unit 7
Discussion Questions:
1. Formulate this case as an assignment problem by constructing the
appropriate cost table, including identifying the corresponding assignee
and tasks.
2. Obtain an optimal solution.
References:
Operations Research
Unit 8
Unit 8
Structure:
8.1 Introduction
Objectives
8.2 Types of Integer Programming Problems
8.3 Gomorys All-IPP Method
Construction of Gomorys constraints
8.4 All IPP Algorithm
8.5 Branch and Bound Technique
Branch and bound algorithm
8.6 Summary
8.7 Glossary
8.8 Terminal Questions
8.9 Answers
8.10 Case Study
8.1 Introduction
In the previous unit, you learnt about mathematical formulation of the
problem and Hungarian method algorithm. You also learnt the routing
problem and the travelling salesman problem. In this unit you will study
about Integer Programming Problem (IPP). The IPP is a special case of
Linear Programming Problem (LPP), where all or some variables are
constrained to assume non-negative integer values. In LPP, the decision
variables as well as slack or surplus variables were allowed to take any real
or fractional value. However, there are certain real life problems in which the
fractional value of the decision variables has no significance. For example, it
does not make sense saying 1.5 men working on a project or 1.6 machines
in a workshop. The integer solution to a problem can, however, be obtained
by rounding off the optimum value of the variables to the nearest integer
value. This approach can be easy in terms of economy of effort, time and
cost that might be required to derive an integer solution but this solution may
not satisfy all the given constraints. Secondly, the value of the objective
function so obtained may not be an optimal value. Integer programming
techniques come to our rescue during such scenarios.
Operations Research
Unit 8
Integer LP problems are those in which some or all of the variables are
restricted to integer (or discrete) values. An integer LP problem has
important applications. Capital budgeting, construction scheduling, plant
location and size, routing and shipping schedule, batch size, capacity
expansion, fixed charge, etc are few problems which demonstrate the areas
of application of integer programming.
In this unit, you will learn about Integer Programming Problem (IPP) and the
Gomorys all-IPP method. You will also learn all IPP algorithm and the
branch and bound technique.
Objectives:
After studying this unit, you should be able to:
analyse an Integer Programming Problem (IPP)
solve an IPP using Gomorys method
describe the branch and bound technique
j = 1, 2, ,n
Operations Research
Unit 8
If all the variables are forced to take only integral value that is k = n, it is
called an all (or pure) integer programming problem. If some of the variables
are restricted to take integral value and the remaining (n k) variables take
any non-negative value, then the problem is known as a mixed integer
programming problem.
Self Assessment Questions
1. Integer programming is applied to problems that involve discrete
variables. (True/False)
2. If some variables take on non-negative values, then it is known as pure
IPP. (True/False)
Operations Research
Unit 8
yB
xB
Y1
y2
y3
y4
y2
y10
y11
y12
y13
y14
y3
y20
y21
y22
y23
y24
y00
y01
y02
y03
y04
Operations Research
Unit 8
fig x j f10 or
j 1, 4
or
fij. x j f10
j 1, 4
j 1,4
Where Gsla (1) is a slack variable in the above first Gomorys constraint.
The additional constraint to be included in the given LPP is towards
obtaining an optimum all integer solution. After adding the constraint, the
optimum simplex table is depicted as in table 8.2.
Table 8.2: Optimum Simplex Table
yB
xB
y1
y2
y3
y4
Gsla(1)
y1
y10
y11
y12
y13
y14
y2
y20
y21
y22
y23
y24
Gsla(1)
f10
f11
f14
y00
y01
y02
y03
y04
y05
Since f10 is negative, the optimal solution is unfeasible. Thus the dual
simplex method is to be applied for obtaining an optimum feasible solution.
After obtaining this solution, the above referred procedure is applied for
constructing second the Gomorys constraint. The process is to be
continued till all the integer solution has been obtained.
Self Assessment Questions
3. An optimum solution to IPP is first obtained by using _____________.
4. With the addition of Gomorys constraint, the problem is solved by
____________________.
Operations Research
Unit 8
Operations Research
Unit 8
Step 1: Convert the minimisation IPP into maximisation form. Ignore the
integrality condition.
Step 2: Introduce the slack or surplus variables, if needed to convert the
inequations into equations and obtain the optimum solution of the given LPP
by using simplex algorithm.
Step 3: Test the integrality of the optimum solution:
a) If the optimum solution contains all integer values, an optimum basic
feasible integer solution has been obtained.
b) If the optimum solution does not include all integer values, then proceed
to the next step.
Step 4: Examine the constraint equations corresponding to the current
optimum solution. Let these equations be represented by:
n1
j0
y ij , x j b i , (i 0, 1, 2, ........, m1 )
j0
f kj , x j fko
G sla (1) f ko
n1
j0
f kj . x j
Operations Research
Unit 8
Step 7: Start with a new set of equation constraints. Find the new optimum
solution by dual simplex algorithm, that is, choose a variable to enter into
the new solution having the smallest ratio:
{(Cj Zj)/ yij; yij<0} so that Gsla (1) is the initial leaving basic variable.
Step 8: If the new optimum solution for the modified LPP is an integer
solution, it is also feasible and optimum for the given IPP. If it is not an
integer solution, then return to step 4 and repeat the process until an
optimum feasible integer solution is obtained.
Solved Problem: 1
Find the optimum integer solution of the following all IPP.
Maximise
z = x 1 + x2
Subject to the constraints
3x1 +2 x2 5;
x2 2
x1, x2 > 0 and are integers
Solution: Step 1: Introduce the slack variables s1>0 and s2 > 0, using
simplex method an optimum noninteger solution is depicted in the
table 8.3:
Initial iteration. Non- integer optimum solution
Table 8.3: Optimal Non-integer Solution
cJ -->
CB
1
1
Z = 7/2
1
Variable in
Basis B
X1
X2
Solution Values
b (= xB)
1/3
2
cj-zj
X1
X2
S1
S2
1
0
0
0
1
0
1/3
0
-1/3
-2/3
1
-1/3
Here, since all cj-zj 0, the optimal solution is : x1=1/3 and x2= 2 and
max Z = 7/2
Step 2: Since x1 is the only basic variable whose value is a non-negative
fraction, we consider the first row for generating the Gomory Cut.
Considering x1 equation as the source row we write:
1/3 = x1+ 0 x2 + 1/3 s1 2/3 s2 (x1- source row)
Operations Research
Unit 8
cJ -->
CB
Variable in
Basis B
Solution
Values b
(= xB)
X1
X2
S1
S2
Sg1
X1
1/3
1/3
-2/3
X2
Sg1
-1/3
-1/3
-1/3
Z = 7/2cj-zj
Ratio: Min(cj-zj) / y3J(<0)
0
-
0
-
-1/3
1
-1/3
1
0
-
Step 4: Since solution depicted in table 8.4 is infeasible, apply the dual
simplex method to find feasible as well as an optimal solution. The key
row and key columns are marked in table 8.4. The new solution is
obtained by applying following row operations and is depicted in
table 8.5.
Operations Research
Unit 8
cJ -->
CB
1
1
0
X1
X2
S1
S2
Sg1
1
0
0
0
0
1
0
0
0
0
1
0
-1
1
1
0
1
0
-3
-1
cj xj
_______________________ (1)
j1
a ij x j b i
i 1, 2,...., m
___________________ (2)
j1
Operations Research
xj is integer valued,
Unit 8
j = 1, 2, ., r ______________________ (5)
This is the main idea behind the branch and bound technique.
Consider any variable xj, and let k be some integer value satisfying
LjkUj 1. Then clearly an optimum solution (1) through (5) also satisfies
either linear constraint.
xj k + 1
_________________________________ (6)
Operations Research
Unit 8
Operations Research
Unit 8
Step 2: Since the solution is non-integral, let us choose x1, i.e., x1* = 9/2
being the largest fractional value.
Step 3: Considering the value of z as initial upper bound, i.e., z = 63; the
lower bound is obtained by rounding off the values of x1 and x2 to
the nearest integers, i.e., x1 = 4 and x2 = 3. Then the lower bound
is z = 55.
Step 4: Since [x1*] = [9/2] = 4; We have,
Sub-Problem 1: Maximise Z = 7x1 + 9x2 subject to:
-x1 + 3x2 6, 7x1 + x2 35, 0 x1 4, 0 x2 7.
Sub-Problem 2: Maximise Z = 7x1 + 9x2 subject to:
-x1 + 3x2 6, 7x1 + x2 35, x1 5, 0 x2 7.
Step 5: Optimum solutions for the sub-problems are obtained as follows:
Sub-Problem 1: x1= 4, x2=10/3 and maximum z = 58.
Sub-Problem 2: x1= 5, x2=0 and maximum z = 35.
Since the solution for sub-problem 1 is not in integers, we
subdivide it into the following two sub-problems:
Sub-Problem 3:
Maximise Z = 7x1 + 9x2
subject to
-x1 + 3x2 6,
7x1 + x2 35 ,
and 0 x1 4, 0 x2 3.
Sub-Problem 4:
subject to
-x1 + 3x2 6,
7x1 + x2 35,
and 0 x1 4, x2 4.
Step 6: The optimum solutions for the sub-problem 3 and 4 are:
Sub-Problem 3: x1 = 4, and x2= 3 and maximum z=55.
Sub-Problem 4: No feasible solution.
Step 7: Among the recorded integer valued solutions, since the largest
value of z is 55; the required optimum solution is v1 = 4, x2 = 3
and maximum z = 55.
Operations Research
Unit 8
The entire branch and bound procedure for the given problem is depicted in
figure 8.2:
Fig. 8.2: The Flow of Sub-Problems in the Branch and Bound Problems
8.6 Summary
Let us recapitulate the important concepts discussed in this unit:
Linear integer programming problems can be classified into pure, mixed
integer and zero- one interger programming problems.
Gomorys constraint represents the necessary conditions for integrability
and eliminates some non-integer solution without losing any integral
solution
The most general technique for a solution of such constrained
optimisation problems is the branch and bound technique.
This unit also describes two algorithms to determine the optimal solution
for an integer programming problem. One is the All IPP algorithm and
the other is the branch and bound algorithm.
Sikkim Manipal University
Operations Research
Unit 8
8.7 Glossary
Non-basic variables: variables set to value zero in order to solve the value
of basic variables.
Equation constraints: A constraint written in the form of an equation
containing a =.
8.9 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
True
False
Simplex method
Dual simplex method
True
False
True
False
Terminal Questions
1. At the end of the 8th iteration we get the optional solution to the IPP is x1
= x2 = 0, x3 = 1, z* = 13. For more details refer Section 8.5
2. An integer programming problem(IPP) can be described as follows:
Determine the value of unknowns x1, x2, , xn
Sikkim Manipal University
Operations Research
Unit 8
j = 1, 2, ,n
Operations Research
Unit 8
upon business during the last month and realised that over 80 percent of the
orders for kitchen sets could not be filled because items needed to complete
the sets were not in stock at the local warehouse. She also realised that
Wood World was losing customer goodwill and business because of stock
outs. The furniture mega store was gaining a reputation for slow service and
delayed deliveries causing customers to turn to small competitors that sold
furniture directly from the showroom floor. Annie decided to investigate the
inventory situation at the local warehouse. She walked the short distance to
the building next door and gasped when she stepped inside the warehouse.
What she saw could only be described as chaos. Spaces allocated for some
items were overflowing into the aisles of the warehouse while other spaces
were completely bare. She walked over to one of the spaces overflowing
with inventory to discover the item that was overstocked. She could not
believe her eyes! The warehouse had at least 30 rolls of pea-green
wallpaper! No customer had ordered pea-green wallpaper since 1973! Annie
marched over to Don demanding an explanation. Don said that the
warehouse had been in such a chaotic state since his arrival one year ago.
He said the inventory problems occurred because management had a policy
of stocking every furniture item on the showroom floor in the local
warehouse. Management only replenished inventory every three months,
and when inventory was replenished, management ordered every item
regardless of if it had been sold. Don also said that he had tried to make
management aware of the problems with overstocking unpopular items and
under stocking popular items, but that management would not listen to him
because he was simply an expeditor.
Annie understood that Wood World required a new inventory policy. Not
only was the mega store losing money by making customers unhappy with
delivery delays, but it was also losing money by wasting warehouse space.
By changing the inventory policy to stock only popular items and replenish
them immediately when they are sold, Wood World would ensure that the
majority of customers receive their furniture immediately and that the
valuable warehouse space was utilised effectively.
Annie needed to sell her inventory policy to the management. Using her
extensive sales experience, she decided that the most effective sales
strategy would be to use her kitchen department as a model for the new
inventory policy. She would identify all kitchen sets comprising 85 percent of
Sikkim Manipal University
Operations Research
Unit 8
Wallpaper
Light Fixtures
Cabinets
Operations Research
(T3) White
checkered tile with
blue trim doors
Unit 8
Countertops
Dishwashers
Sinks
Ranges
(D1) White
energy- saving
dishwasher
(R1) White
electric oven
(D2) Ivory
energy- saving
dishwasher
Discussion Questions:
1. Formulate and solve an IPP model to maximise the total number of
kitchen sets (and thus the number of customer orders) Wood World
stocks in the local warehouse. Assume that when a customer orders a
kitchen set, all the particular items composing that kitchen set are
replenished at the local warehouse immediately.
2. How many of each feature and style should Wood World stock in the
local warehouse? How many different kitchen sets are in stock?
3. Annie convinces the management that the kitchen department should
serve as a testing ground for future inventory policies. To provide
adequate space for testing, management decides to allocate all the
space freed by the nursery department to the kitchen department. The
Sikkim Manipal University
Operations Research
Unit 8
extra space means that the kitchen department can store not only the
dishwashers and ranges from part (c), but also all sinks, all countertops,
three of the four light fixtures and three of the four cabinets. How much
does the additional space help?
References:
Kanti Swarup & Gupta P. K., & Hira D. S., & Manmohan (2004).
Operation Research. Sultan Chand and Sons.
Operations Research
Unit 9
Unit 9
Structure:
9.1 Introduction
Objectives
9.2 Queuing Theory
9.3 Operating Characteristics of a Queuing System
9.4 Constituents of a Queuing System
Arrival pattern
Completely random arrivals
9.5 Service Facility
9.6 Queue Discipline
Customer behaviour
Server behaviour
9.7 Summary
9.8 Glossary
9.9 Terminal Questions
9.10 Answers
9.11 Case Study
9.1 Introduction
In the previous unit, you learnt the Integer Programming Problem (IPP) and
the Gomorys all-IPP method. You also learnt all IPP algorithm and the
branch and bound technique. This chapter introduces the idea of queuing
theory. The formation of waiting lines or queues is common and is usually
formed by elements, people or events that are awaiting some form of
service. Who isnt familiar with queues in our daily existence? On the way to
your workplace, you have to wait at the bus stop, wait for the traffic light to
turn green, wait for your turn to enter the lift. You come across cars in line at
petrol pumps for service, queues to deposit cash at the bank, mobile
subscriber waiting for a new connection, customers waiting at the checkout
counter, goods in production/warehouse waiting to be shipped and even
aircrafts waiting for a free runway to take off.
The one thing common to all the above mentioned examples is that
customers arrive at a service centre and wait for their turn to receive the
service. Though the arrival of customers is irregular and the time taken for
Sikkim Manipal University
Operations Research
Unit 9
the service is not consistent, queues build up during hours of demand and
disappear during the lull period.
In commercial or industrial situations, it may not be economical to have
waiting lines. On the other hand, it may not be feasible or economical to
totally avoid queues. An executive dealing with the system then would like to
find the optimal facilities to be provided. In this unit, you will learn about
queuing theory, operating characteristics and constituents of a queuing
system. You will also learn about service facility and queue discipline.
Objectives:
After studying this unit, you should be able to:
describe the queuing process
evaluate the Queuing Theory
differentiate the customers behaviour.
Operations Research
Unit 9
For example
Supermarkets must decide how many cash register checkout positions
should be opened.
Gasoline stations must decide how many pumps should be opened and
how many attendants should be on duty.
Manufacturing plants must determine the optimal number of mechanics
to have on duty in each shift to repair machines that break down.
Banks must decide how many teller windows to keep open to serve
customers during various hours of the day.
Self Assessment Questions
1. Customers arrive at a bank at regular intervals. (True/False)
2. Queuing identifies the optimal service facilities to be provided.
(True/False)
3. Queuing theory is based on the deterministic model. (True/False)
Queue length: The number of customers in the waiting line reflects one
of the two conditions. Short queues could mean either good customer
service or too much capacity. Similarly, long queues could indicate
either low server efficiency or the need to increase capacity
Waiting time in queue: Long lines do not reflect long waiting times if
the service rate is fast. However, when waiting time seems long to
customers, they perceive that the quality of service is poor. Long waiting
times may indicate a need to adjust the service rate of the system or
change the arrival rate of customers.
Operations Research
Unit 9
Waiting time in system: The total elapsed time from entry into the
system until exit from the system may indicate problems with customers,
server efficiency or capacity. If some customers are spending too much
time in the service system, there may be a need to change the priority
discipline, increase productivity or adjust capacity in some way.
Operations Research
Unit 9
Operations Research
Unit 9
Operations Research
Unit 9
Operations Research
Unit 9
Situation
Arriving
Customers
Service Facility
1.
Shoppers
Checkout counters
2.
Automobiles
Road Network
3.
Electronic
messages
Transmission lines
4.
Banking transactions
Bank patrons
Bank tellers
5.
Computer
Programmers
Central processing
unit
6.
Theatre visitors
Ticket booking
windows
7.
Trucks
8.
Registration of unemployed at
employment exchange
Unemployed
personnel
Registration
assistants
9.
Occurrences of fires
Fires
Firemen and
equipment
10.
Ships
Harbour and
docking facilities
11
Service calls
Policemen
Operations Research
Unit 9
I. First In First Out (FIFO) If the customers are served in the order of
their arrival. E.g., Toll road
II. Last In First Out (LIFO) Where the last in items are operated first.
E.g., Cargo handling situations, getting down from the bus, etc.
Dynamic queuing disciplines are based on the individual customer attributes
in the queue:
I. Service in Random Order (SIRO): Here customers are selected for
service at random irrespective of their arrivals in the service system.
II. Priority service: Under this rule, customers are grouped in priority
classes on the basis of some attributes such as service time or
urgency, and FIFO rule is used within each class to provide service.
E.g., premium queues in temples like Tirupati.
III. Pre-emptive priority (or Emergency): Under this rule, an important
customer is allowed to enter into the service immediately after entering
into the system even if a customer with lower priority is already in
service. E.g., Emergency service in hospitals, ambulance and fire
brigades in traffic signals
9.6.1 Customer behaviour
a) Balking: Arriving customers are said to balk if they do not join a queue
because of their reluctance to wait or some customers even before
joining the queue get discouraged by seeing the number of customers
already in service system or estimating the excessive waiting time for a
desired service, decide to return for service at a later time. In queuing
theory this is known as balking.
b) Collusion: Customers may be in collusion, meaning that only one
person would join the queue, but demand service on behalf of several
customers.
c) Reneging: Impatient customers who would not wait beyond a certain
time and leave the queue are said to renege.
For example, a customer who has just arrived at a grocery store and
finds that the salesmen are busy in serving the customers already in the
system, will either wait for service till his patience is exhausted or
estimates that his waiting time may be excessive and so leaves
immediately to seek service elsewhere.
Sikkim Manipal University
Operations Research
Unit 9
9.7 Summary
Let us recapitulate the important concepts discussed in this unit:
The waiting line theory or queuing theory analysis suggests the number
of facilities required and the cost of customers waiting time and the
optimum service level.
The queuing theory contributes vital information required for balancing
the cost of service and cost associated with waiting time of the
customer.
A queuing system is said to be in transient state when its operating
characteristics are dependent upon time.
9.8 Glossary
Erlang family of service time distributions: two constraint generalization
of exponential distribution.
Statistical pattern: set of probability distributions
Operations Research
Unit 9
9.10 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
False
True
False
Queue length
Long waiting
Collective utilisation
True
True
Series, parallel
Constant, varying
Jockeying
Balking, Collusion, reneging, jockeying
Terminal Questions
1. The operating characteristics of a queuing system are queuing length,
number of customers in system,waiting time in queue, waiting time in
system and service facility utilisation. For more details, refer section 9.3
2. Size of population, pattern of arrivals and behavior of arrivals. For more
details refer section 9.4.1
3. Static queuing disciplines and dynamic queuing disciplines. For more
details refer section 9.5
Operations Research
Unit 9
Operations Research
Unit 9
References:
E-References:
http://www.mrt.ac.lk/maths/TMJAC/Operational%20Research
%20Techinques%20II/OR%20Lecture%203.pdf, accessed on Nov 9,
2009
Operations Research
Unit 10
Unit 10
Structure:
10.1 Introduction
Objectives
10.2 Mathematical Analysis of Queuing Process
10.3 Properties of Queuing System
Arrivals
Service
Queuing discipline
10.4 Notations
10.5 Service System
10.6 Single Channel Models
10.7 Multiple Service Channels
10.8 Erlang Family of Distribution of Service Times
10.9 Applications of Queuing Theory
10.10 Limitations of Queuing Theory
10.11 Summary
10.12 Glossary
10.13 Terminal Questions
10.14 Answers
10.15 Case Study
10.1 Introduction
In the previous unit, you learnt queuing theory, operating characteristics and
constituents of a queuing system. You also learnt about service facility and
queue discipline. Queuing theory is a collection of mathematical models of
various queuing systems that takes in input parameters and provides
quantitative parameters describing the system performance.
Because of the random nature of the processes involved, the queuing
theory is rather demanding and all models are based on very strong
assumptions (not always satisfied in practice). Many systems (especially
queuing networks) are not soluble at all, so the only technique that may be
applied is simulation.
Operations Research
Unit 10
Operations Research
Unit 10
The probability distribution of the arrivals will then be different from that of
the initial state. The state, in which the probability distribution remains the
same, is called the steady state and the system is said to have acquired a
state of statistical equilibrium.
In the steady state, there will be variations in the queue from time to time but
the probability distributions representing the queuing process will remain the
same and are independent of the time at which the system is examined.
(t)x/ x !
Where x = 0, 1, 2 ..
10.3.2 Service
Single serve channel/multiple channels
Single queue/infinite capacity
10.3.3 Queuing discipline
First come first serve
Note: When the number of arrivals follows Poisson distribution (discrete),
the inter-arrival time, that is, the time between arrivals follows exponential
distribution (continuous).
Operations Research
Unit 10
10.4 Notations
In the previous section, you learnt the properties of the system. You will now
learn about the notations. The queuing systems with which you are
concerned are denoted by M/M/1 and M/M/c where M stands for exponential
inter arrival and exponential service time distributions and the third character
indicates the number of channels (or servers) available (1 or c).
Table 10.1 depicts a list of notations that are commonly used.
Table 10.1: List of Notations Used
Notation
Z
H
p
c
m
n
Pn
E(m)
E(m/m > 0)
E(n)
W
E(w)
P(w = 0)
E(w/w > 0)
V
F(v)
E(v)
Meaning
= Average number of arrivals per time
= Average number of customers served per unit time
= Traffic intensity
= Number of service channels
= Number of customers in the queue or the length of the queue
= Number of customers in the System (Number of customers in the
queue + Number of customers being served )
= Steady state probability of finding n people in the system
= Average length of queue
= Average length of a non-empty queue
= Average number of customers in the system
= Steady state waiting time of a customer
= Probability of not having to waiting time of a customer
= Average waiting time of a customer given that there is a queue
= Time spent by a customer in the system in steady state
= Waiting time in the queue + Service time
= Probability density function of the time spent by a customer in the
system
= Average time spent by a customer in the system
Operations Research
Unit 10
ii. Single server Several queues In this type of facility there are several
queues and the customer may join any one of these but there is only one
service channel. Figure 10.2 depicts the single server several queues
model.
Sikkim Manipal University
Operations Research
Unit 10
Queues
Service facility
Customer leave
Arrivals
iii. Several (parallel) servers Single queue In this type of model there
is more than one server and each server provides the same type of facility.
The customers wait in a single queue until one of the service channels is
ready to take them in for servicing. Figure 10.3 depicts the several (parallel)
servers single queue model.
Service stations
Fig. 10.3: Several (Parallel) Servers Single Queue Model
Operations Research
Unit 10
The service time indicates the amount of time needed to service a single
customer.
Service rates and times are reciprocal of each other and either of them
is sufficient to indicate the capacity of the facility.
Operations Research
Unit 10
Pn 1
2
( )
E ( m / m 0)
E ( n)
E ( w)
( )
E ( m)
1
P ( w o) 1
1
E (v )
f ( w) (1 ) e w ( )
f (v ) ( ) e v ( )
E ( w / w 0)
Fig. 10.6: Formulae for Poisson Arrivals, Exponential Service, Single Channel
Queuing Models Infinite Population
Sikkim Manipal University
Operations Research
Unit 10
Fig. 10.7: Formulae for Poisson Arrivals, Exponential Service, Single Channel
Queuing Models Number of Customers Linked to N
Solved problem 1:
Patrons arrive at a small post office at the rate of 30 per hour. Service by the
clerk on duty takes an average of 1 minute per customer.
a) Calculate the mean customer time:
i) Spent waiting in line
ii) Spent receiving or waiting for service
b) Find the mean number of persons:
i) In line
ii) Receiving or waiting for service
Solution:
Mean arrival rate = 30 customers per hour
=
Traffic intensity
(a) (i) Mean customer time spent waiting in line
Operations Research
Unit 10
= 2 minutes
(b) (i) Mean number of persons in line
= 0.5 customer
(ii) Mean number of persons receiving or waiting for service
Customer
Solved problem 2:
The Tool Companys quality control department is manned by a single clerk,
who takes an average of 5 minutes in checking parts of each of the
machines coming for inspection. The machines arrive once in every 8
minutes on an average. One hour of machine is valued at Rs. 15 and a
clerks time is valued at Rs. 4 per hour. What is the average hourly queuing
system costs associated with the quality control department?
Solution:
Mean arrival rate l = 1/8 per minute =
Mean service rate
per hour
hours
An average arrival of
Operations Research
Unit 10
Solved problem 3:
A television repairman finds that the time spent on his jobs has an
exponential distribution with a mean of 30 minutes. If he repairs sets in the
order in which they came in and if the arrival of sets follow a Poisson
distribution with an average rate of 10 per 8 hour day, what is the
repairmans expected idle time each day? How many jobs are ahead of the
average set just brought in?
Solution:
= 10/8 = 5/4 sets/hr;
= (1/30) 60 = 2 sets/hr.
Number of hours for which the repairman remains busy in an 8-hour day =
8* ( / ) = 8(5/8) = 5 hrs.
Repairmans expected idle time in an 8 hour day = 8-5 = 3 hrs.
Expected number of T.V sets in the system = (Ls) = / ( - )
= (5/4)/ [2-(5/4)]
= 5/3
= 2 sets (approx.)
Solved problem 4:
Arrivals at a telephone booth are considered to follow Poisson distribution
with an average time of 10 minutes between one arrival and the next. The
length of phone call is assumed to be distributed exponentially, with a mean
of 3 minutes.
a) What is the probability that a person arriving at the booth will have to
wait?
b) The telephone department will install a second booth when convinced
that an arrival would expect to wait for at least 3 minutes for a phone
call. By how much should the flow of arrivals increase in order to justify
a second booth?
c) What is the average length of the queue that forms from time to time?
d) What is the probability that it will take him/her more than 10 minutes
altogether to wait for the phone and complete his call.
Operations Research
Unit 10
Solution:
= 1/10 person/minute;
= 1/3 = 0.33 person/minute.
a) Probability that a person arriving at the booth will have to
wait = 1- P0 = 0.3
b) The installation of the 2nd booth will be justified only if the arrival rate is
more than the waiting time.
Let be the increased arrival rate. Then expected waiting time in the
queue will be (Wq) = / ( - )
= / 0.33(0.33- )
Or =0.16
Hence the increase in the arrival rate is 0.16-0.10 = 0.06 arrivals
/minute
c) Average length of non-empty queue = / ( - ). = 0.33/0.23 = 1.42857
Customers (approx.)
d) Probability of waiting for 10 minutes or more = P (t>=10)
-( - )t
= ( / ) ( - ) e
dt
= 5 per hour
=8
Sikkim Manipal University
Operations Research
Unit 10
Pr =
n
n
P0 for 0
The probability distribution for the number of patients either waiting for
treatment or receiving treatment is as shown in table 10.2.
Table 10.2: Probability Distribution
(b)
Pn
0.271
0.217
0.173
0.139
0.111
0.089
(1) The mean value for the number of patients in the emergency room
= 0 x (0.271) + 1 x (0.217) + 2 x (0.173) + 3 x (0.139) + 4 x (0.111) + 5
x (0.089)
= 1.868332032.
Solved problem 6:
Cars arrive at a toll gate on a frequency according to Poisson distribution
with mean 90 per hour. Average time for passing through the gate is 38
seconds.
Drivers complain of long waiting time. Authorities are willing to decrease the
passing time through the gate to 30 seconds by introducing new automatic
devices. This can be justified, if under the old system, the number of waiting
cars exceeds 5 and the percentage of gates idle time under the new system
should not exceed 10.
In the above scenario, can the new device be justified?
Sikkim Manipal University
Operations Research
Unit 10
Solution:
In the old system,
Meaning arrival rate = 90 per hour = 1.5 per minute
Mean service rate =
Traffic intensity =
per minute
=
Operations Research
Unit 10
d) How many spaces should be provided in front of the window so that all
the arriving customers can wait in front of the window at least 20% of
the time?
Solution:
(a) = 10 per hour;
10/12 = 5/6 per hour
(b) p = / = 10/12 = 5/6
The probability that an arriving customer can drive directly to the
space in front of the window
= P0+P1+P2
= (1-p)(1 + p +p2 )
=
(c) Probability that an arriving customer has to wait outside the indicated
space = probability that there are at least 3 customers in the space in
front of the window
=
0 + p1 + p2 + p3)
=1 (1- p) (1 + p + p2 + p3)
= .48
(d) The average waiting time of a customer in queue.
0 + p1 + p2 0.20
0 + p1 + p2 = 0.42
As even with one space in front of the window, 42 percent of the times an
arrival waits in the space, the number of spaces required is one or more.
Self Assessment Questions
4. The expected number of customers in non-empty queue is given by
___________.
Sikkim Manipal University
Operations Research
Unit 10
Fig. 10.8: Formulae for Poisson Arrivals, Exponential Service Multi channel
Queuing Models Infinite Population
Solved problem 8:
Ships arrive at a port at a rate of one in every 3 hours, with a negative
exponential distribution of interarrival times. The time a ship occupies a
Sikkim Manipal University
Operations Research
Unit 10
berth for unloading and loading has a negative exponential distribution with
an average of 12 hours. If the average delay of ships waiting for berths is to
be kept below 6 hours, how many berths should there be present at the
port?
Solution:
Pe 1/( c 1)! / ( c p ) 2 1
rp
= 6.65
This is greater than 6 hours and inadequate.
When
P0
Operations Research
Unit 10
Arrival of customers follows Poisson distribution with mean 8 per hour for
counter 1 and 5 per hour for counter 2.
a) What are the average waiting times per customer of each counter?
b) If each counter could handle all withdrawals irrespective of their value
how would the average waiting time change?
Solution:
(a) For counter one
1=
(b) When each counter can handle all withdrawals irrespective of their
value with an average arrival rate of 13 per hour, mean service time
of 10 per hour at each counter and number of channels = 2,
= 4. 71P0 = 0.21
Average waiting time E(w)
= P0 = 4.38961039 minutes
Self Assessment Questions
6. Write the formula for calculating the expected number of customers in
the system.
7. Write the formula for calculating the average waiting time of the
customer in the queue.
Sikkim Manipal University
Operations Research
Unit 10
Operations Research
Unit 10
Operations Research
Unit 10
= 6 minutes
Operations Research
Unit 10
10.11 Summary
Let us recapitulate the important concepts discussed in this unit:
This unit discusses about the mathematical analysis of Queuing process &
how the service system is useful in different process.
10.12 Glossary
Queuing discipline: It refers to the decision rule by which the customers
are selected from queue for service
Operations Research
Unit 10
10.14 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
k 1
2
2k ( )
9.
k 1
2k ( )
Terminal Questions
1.
2.
3.
4.
Refer 10.8
i) 12.4 min ii) 0.79= one stenograph iii) 22.4 min.
Pn=1/4(1/4)n, Ls = 3 customers , W s= 6 min.
Refer 10.5
Operations Research
Unit 10
Operations Research
Unit 10
Operations Research
Unit 10
Operations Research
Unit 11
Unit 11
Structure:
11.1 Introduction
Objectives
11.2 Finite Queuing Models
Finite queuing tables
Measures of system efficiency
Use of finite queuing tables
11.3 Summary
11.4 Glossary
11.5 Terminal Questions
11.6 Answers
11.7 Case Study
11.1 Introduction
In the previous unit, you learnt about mathematical analysis of queuing
theory, notations and service systems involved in queuing theory. You also
learnt about Erlang family of distribution of service times and the
applications and limitations of queuing theory. The models discussed so far
relate to situations involving infinite population of customers, that is, the
queue can increase indefinitely. However, you may come across scenarios
where the possible number of arrivals is limited and small. In this unit, we
will discuss such scenarios.
In a production shop, if the machines are considered as customers requiring
service from repair crews or operators, the population is restricted to the
total number of machines in the shop. In a hospital ward, the probability of
the doctors or nurses being called for service is governed by the number of
beds in the ward. Similarly, in an aircraft the number of seats is finite and
the number of stewardesses provided by the airlines will be based on the
consideration of the maximum number of passengers who can demand
service.
SikkimManipalUniversity
Operations Research
Unit 11
In case of a queuing system with infinite population, you can improve the
efficiency of the system in terms of reduced average length of queues,
average waiting time and time spent by the customer in the system. To do
this, you need to increase the number of service channels. However, such
increases mean additional cost and will have to be balanced with the
benefits likely to accrue. If the queuing system in a machine shop is under
study, the cost of providing additional maintenance crews or operators can
be compared with the value of additional production due to reduced
downtime of the machines. Sometimes it is not possible to quantify the
benefitsthe management will have to base its decisions on the desired
standards for customer service. In this unit, you will learn the finite queuing
models, finite queuing tables and the use of these models and tables.
Objectives:
After studying this unit, you should be able to:
describe queuing models and various queuing disciplines
interpret and apply finite queuing tables to solve problems
Analysis of the relevant queuing models allows you to identify the cause of a
queuing issue and assess the impact of the proposed changes. Typically,
you need to analyse a queuing model to minimise the sum of:
Customer waiting cost
Service capacity cost
SikkimManipalUniversity
Operations Research
Unit 11
SikkimManipalUniversity
Operations Research
Unit 11
Random:
This principle states that the system may randomly select a customer for
service. For example, in a machine shop if a single operator is attending to
several machines and several machines call for his attention at a time, he
may attend first to the one nearest to him.
Last Come-First Served (LCFS):
This discipline is based on the stack method.
Most quantitative parameters, such as the average queue length and
average time spent in the systemdo not depend on priority queues. That is
the reason most models either do not take the queuing discipline into
account at all or assume the normal FCFS queue. The only parameter that
depends on the queuing discipline is the variance (or standard deviation) of
the waiting time.
The analysis of finite queuing models is more complex than those with
infinite population although the approach is similar.
Simple waiting line model
The simplest waiting line model assumes that arrivals join a queue that is of
unlimited size, waiting in line until their turn for service comes on a first-cumfirst-serve basis and then enter a service facility consisting of a single
channel.
If, W = Average time spent in queue (i.e., sum of the expected waiting time
and expected service time) = Average rate of units passing through the
system per unit tim e L = Average number of units in the system.
Then, L= W
If = Average service rate
1/ = Expected inter arrival time
1/ = Expected service time
Further, the utilisation factor, (i.e., the expected fraction of time the server
is busy) is given by:
= / s
Where, (s) is the fraction of the system's service capacity that being
utilised on the average by arriving customers (C)
SikkimManipalUniversity
Operations Research
Unit 11
It is assumed that machines are kept running (or in operation) all the time,
except when in repairs or waiting for repair crew to attend. If M repair crews
are available, the time taken by any crew follows a negative exponential
distribution with mean T. Naturally, a machine which has broken down will
have to wait for repairs if all the repair crews are busy.
11.2.2 Measures of system efficiency
For a given set of machines, the efficiency of the repair system may be
judged by the extent to which machines have to wait for repairs. If W is the
average time for which a machine has to wait, the efficiency factor F is
defined as,
SikkimManipalUniversity
Operations Research
Unit 11
T U
T U W
NT
FNX
TWU
NW
L
N (1 F)
T W U
NU
J
NF (1 X)
T W U
HJ
TU
F
HLJ T U W
H
SikkimManipalUniversity
Operations Research
Unit 11
Source: Shim, K. Jae, & Siegel, G. Joel. Operations Management . Barron's Educational Series
Operations Research
Unit 11
The overall efficiency F of the system will increase with the number of
service channels (M) provided. As mentioned earlier, addition of service
crews involves cost, which should be justified by the increase in the
efficiency of the system that is additional running time of machines.
However, it will be seen from the table that as M increases the rate of
increase in efficiency decreases. The practical significance is that beyond a
certain value of M, it is not worthwhile increasing M as there would be no
appreciable increase in the efficiency of the system.
SikkimManipalUniversity
Operations Research
Unit 11
Problem 1:
In a chemical factory, there are five hoppers of identical size which feed
material to grinding mills. Due to changes in the requirement of material,
there are variations in the time taken for emptying the hoppers. On the
basis of past experience, it was found to follow negative exponential
distribution with an average of 10 hours between getting emptied.
Whenever a hopper gets empty, it has to be filled by a pay loader.
Although the capacity of the hopper is the same, the time taken to fill the
hoppers varies due to different locations from where the material is to be
loaded.
The time for filling the hopper also was found to follow negative
exponential distribution with an average of 2.5 hours. The company hires
the pay loaders at a cost of Rs. 100 per hour irrespective of whether it is
operated or not. If the mill has to be stopped due to the empty hopper, it
costs Rs. 1000 per hour in terms of loss of profits. To determine the
number of pay loaders which the company should engage to minimise
overall cost you need to create a finite queuing table.
Solved Problem 1:
For the above scenario, determine the number of pay loaders the company
should engage to minimise overall cost.
Solution: Since T = 2.5 and U = 10
X
T
2.5
0.200
T U 2.5 10
0.028
0.998
0.194
0.976
0.689
0.801
SikkimManipalUniversity
Operations Research
Unit 11
2.
0.998
0.976
0.801
3.
N(1 X)
4.00
4.00
4.00
4.
3.992
3.904
3.204
5.
3,992
3,904
3,204
6.
300
200
100
7.
3,692
3,704
3,104
As increasing the number of pay loaders beyond two reduces the profits,
the company should engage only two pay loaders.
Self Assessment Questions
1. When the possible number of arrivals is limited, then we apply infinite
queuing model. (True/False)
2. The queue discipline in a finite queuing process can be random.
(True/False)
T U
3. The efficiency factor for this model is F
. (True/False)
TW
4. __________ principle states that customers are served one at a time
and that the customer that has been waiting the longest is served first.
5. ______ discipline models give priority to rush jobs and important
customers over others.
11.3 Summary
Let us recapitulate the important concepts discussed in this unit:
This unit on finite queuing theory deals with situations where customers
arrive, wait for the service, get the service and leave the system.
Customers may come individually or in groups from large/small
population, at known/variable times, form one or more queues and move
in a certain order to the service station(s)to avail of services whose
speed may be fixed or variable.
SikkimManipalUniversity
Operations Research
Unit 11
Queuing systems are analysed for determining the optimal service level,
where the total cost of providing service and waiting, is minimised.
An increase in the service level increases the cost of providing service,
but reduces the cost of waiting, while a decrease in the service level
induces opposite changes.
11.4 Glossary
Stack: Last in first out collection
Variance: how far the waiting time in a queue is spread out
Trade-off: compromise at the expense of one or more objectives
2.
SikkimManipalUniversity
Operations Research
Unit 11
11.6 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
False
True
False
First Come - First Served (FCFS)
Priority
Terminal Questions
P (the system is idle) = 1- = 1-0.4 = 0.6.
P (n > 2) = 1 P (n 2) = 1 [P (0) + P (1) + P (2)] = 1- 0.936 =
0.064.
iii. P (no customer waiting to be served) = P (0) + P (1) = 0.84.
iv. P (a customer is being served and none is waiting) = P (1) = 0.24.
For more details, refer section 11.2
1. i.
ii.
Operations Research
Unit 11
a Poisson input process at a mean rate of one per hour. The time required
by a crew to unload and/or load a truck has an exponential distribution. The
expected time required by a one man crew would be two hours.
The cost of providing each additional member of the crew is Rs 10 per hour.
The cost that is attributable to having a truck not in use is estimated to be
Rs 40 per hour. Assume that the mean service rate of the crew is
proportional to its size, what should the size be in order to minimise the
expected total cost per hour?
The general routine is that order pickers assemble orders placed by stores
and load them.Oneof the company trucks wait at the single loads, on a
FCFS basis. They then proceed to the store where the order is destined,
unload and return for another order. Because of the many different routes
and distances, and traffic problems at different times of the day, the time
between arrivals of trucks at the dock is randomaveraging 30 minutes. The
loading time also follows an exponential distributionaveraging 15
minutes.Truck drivers are paid Rs. 10 per hour and the crew of two loaders
are each paid Rs. 4 per hour. Truckers have complained about the long
waiting, so a sample was taken showing that truckers did indeed wait for an
average of 32 minutes. The warehouse manager knows that a second truck
dock would probably solve the problem, but the large capital expenditure
plus the disruption of operations during construction are deterrents to the
solution. Tests are made with different crew patterns and it is found that a
crew of 3 can be used to advantage, reducing the loading time to
10 minutes. As management trainee working with the corporation, help the
corporation in finding suitable answers to the following questions:
Discussion Questions:
a. Is the crew of 3 loaders more economical than the crew of 2?
b. What is the probability that an arriving truck will find at least one truck
already in the system?
c. How long is the crew idle?
d. If another truck dock is available with the present operating
characteristics, how will it affect the total expected cost per hour?
SikkimManipalUniversity
Operations Research
Unit 11
References:
E-References:
SikkimManipalUniversity
Operations Research
Unit 12
Unit 12
Simulation
Structure:
12.1 Introduction
Objectives
12.2 Simulation
12.3 Methodology of Simulation
12.4 Basic Concepts
12.5 Simulation Procedure
Allocation of random numbers
Use of random number tables
12.6 Sample Size
12.7 Summary
12.8 Glossary
12.9 Terminal Questions
12.10 Answers
12.11 Case Study
12.1 Introduction
In the previous unit, you learnt the finite queuing models, finite queuing
tables and the use of these models and tables. Simulation is a numerical
technique for conducting experiments that involve certain types of
mathematical and logical relationships necessary to describe the behaviour
and structure of a complex real world system over an extended period of
time. In other words, it is a quantitative technique that utilises a
computerised mathematical model in order to represent actual decision
making under conditions of uncertainty for evaluating alternative courses of
action based upon facts and assumptions. In this unit, you will learn about
simulation, methodology and basic concepts of simulation. You will also
learn simulation procedure and sample size.
A definition of simulation as given by Shannon:
Simulation is the process of defining a model of a real system and
conducting experiments with this model for the purpose of understanding
the behaviour (within the limits imposed by a criterion or a set of criteria) for
the operation of a system.
Sikkim Manipal University
Operations Research
Unit 12
Objectives:
After studying this unit, you should be able to:
define simulation
explain the application of simulations in business problems
describe the use of random number table
12.2 Simulation
Using simulation, an analyst can introduce the constants and variables
related to the problem, set up the possible courses of action and establish
criteria which act as measures of effectiveness. The major reasons for
applying simulation technique to OR problems may be listed as below:
1. It is an appropriate tool to use in solving a problem when experimenting
on the real system:
a) Would be disruptive.
b) Would be too expensive.
c) Does not permit replication events.
d) Does not permit control over key variables.
2. It is desirable tool for solving a business problem when a mathematical
model is:
a) Complex to solve.
b) Beyond the capacity of available personnel.
c) Not detailed enough to provide information on all important decision
variables.
3. The major reasons for adopting simulation in place of other
mathematical techniques are:
a) It may be the only method available, because it is difficult to observe
the actual reality.
b) Without appropriate assumption, it is impossible to develop a
mathematical solution.
c) It may be too expensive to actually observe the system.
d) There may not be sufficient time to allow the system to operate for a
very long time.
4. It provides a trial-and-error movement towards the optimal solution. The
decision maker selects an alternative to experience the effect of the
selection and then improves the selection. In this way, the selection is
adjusted until it approximates the optimal solution.
Sikkim Manipal University
Operations Research
Unit 12
Operations Research
Unit 12
78466
71923
78722
78870
06401
61208
04754
05003
97118
95983
Page No. 230
Operations Research
Unit 12
By following a convention that even digits signify a head (H) and the odd
digits represent a tail (T), the tossing of a coin can be simulated. The
probability of occurrence of the first set of digits is and that of the other
set is also a condition corresponding to the probability of the occurrence
of a head and the probability of occurrence of a tail respectively.
It is immaterial as to which set of five digits should signify a head. The rule
could be that the digits 0, 1, 2, 3 and 4 represent a head and the digits 5, 6,
7, 8 and 9 a tail. It is only necessary to take care that the set of random
numbers allotted to any event matches with its probability of occurrence. For
instance, if youre interested in allotting random numbers to three events A,
B and C with respective probabilities 0.24, 0.36 and 0.40, choose two digit
random numbers from 00 to 99.
The numbers 00 to 23 signify event A, 24 to 59 signify B and 60 to 99 signify
C. The first set of five random digits in the list of random numbers implies
that the outcome of the first toss of 5 coins is as depicted in table 12.2:
Table 12.2: Outcome of First Toss of 5 Coins
Coin
Random Number
Outcome
1
7
T
2
8
H
3
4
H
4
6
H
5
6
H
Number of
Heads
4
3
4
3
1
1
3
4
3
1
Tails
1
2
1
2
4
4
2
1
2
4
Operations Research
Unit 12
10 throws
20 throws
30 throws
40 throws
50 throws
60 throws
70 throws
80 throws
90 throws
100 throws
0
6
11
14
18
19
21
22
24
27
Table 12.5 compares the final results at the end of 100 throws with the
theoretical probabilities.
Operations Research
Unit 12
Estimated
Probability
Theoretical
Probability
0
1
2
3
4
5
0.03
0.21
0.27
0.33
0.12
0.04
0.03
0.16
0.31
0.31
0.16
0.03
It is observed that the results obtained with the large sample of 100, when
compared are more in favour with the theoretical values than with a sample
of ten sets of numbers.
Self Assessment Questions
4. Simulation may be called experimentation in the _________ _______.
5. Random numbers have the property that any number is ____ to occur.
6. The totality of probability assigned to the variable should always be
equal to _______.
Operations Research
Unit 12
Solved problem 1:
A sample of 100 arrivals of customers at a retail sales depot is according to
the following distribution:
Table 12.6: Samples of 100 Arrivals of Customers
Time between Arrivals (min)
Frequency
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
2
6
10
25
20
14
10
7
4
2
Frequency
.5
12
1.0
21
1.5
36
2.0
19
2.5
3.0
Operations Research
Unit 12
Step 3: Using random numbers from table, sample the random time of
arrival and service time for ten sets of random numbers.
Step 4: Tabulate waiting time of arrivals and idle time of servers.
Step 5: Estimate the percentage waiting time of arrivals and percentage idle
time of servers corresponding to the ten samples.
12.5.1 Allocation of random numbers
The tables 12.8, 12.9 and 12.10 depict the allocation of random numbers
based on time between arrivals and service time.
Table 12.8: Allocation of Random Numbers Time Between Arrivals
Time
Frequency
Cumulative
Between
(2)
Arrivals
Frequency
Cumulative
Probability
Random
Number
(3)
(3)/100
Allocated
(4)
(5)
(1)
0.5
0.02
00 to 01
1.0
0.08
02 to 07
1.5
10
18
0.18
08 to 17
2.0
25
43
0.43
18 to 42
2.5
20
63
0.63
43 to 62
3.0
14
77
0.77
63 to 76
3.5
10
87
0.87
77 to 86
4.0
94
0.94
87 to 93
4.5
98
0.98
94 to 97
5.0
100
1.00
98 and 99
Frequency
(2)
12
21
36
19
7
5
Cumulative
Frequency
(3)
12
33
69
88
95
100
Cumulative
Probability
(4)
0.12
0.33
0.69
0.88
0.95
1.00
Random Nos.
Allocated
(5)
00 to 11
12 to 32
33 to 68
69 to 87
88 to 94
95 to 99
Operations Research
Unit 12
Note that the upper bound of random numbers allocated for each value of
the parameter is one less than the corresponding cumulative frequency,
since you have chosen a range of random numbers from 00 to 99
Table 12.10: Arrival and Service Time
Time
Between
Arrivals
Time of
Arrival
Random
No.
1
2
3
4
5
6
7
8
9
10
78
78
06
04
97
71
78
61
05
95
3.5
3.5
1.0
1.0
4.5
3.0
3.5
2.5
1.0
4.5
3.5
7.0
8.0
9.0
13.5
16.5
20.0
22.5
23.5
28.0
54
24
51
45
46
84
58
58
60
24
Service Time
Random No.
Service
Arrival No.
Arrivals
Time
of
Start
Time
of
Finish
Waiting
Time of
Arrival
Idle
Time
of
server
1.5
1.0
1.5
1.5
1.5
2.0
1.5
1.5
1.5
1.0
3.5
7.0
8.0
9.5
13.5
16.5
20.0
22.5
24.0
28.0
5.0
8.0
9.5
11.0
15.0
18.5
21.5
24.0
25.5
29.0
0.5
0.5
-
3.5
2.0
2.5
1.5
1.5
1.0
2.5
1.0
14.5
Total
The service facility is made available at clock time zero and the server has
to be idle for 3.5 minutes, when the service for first arrival starts. The service
is completed at 5.0 minutes and again the server is idle for 2 minutes till the
second arrival joins the system. The first three arrivals get immediate
service and they dont have to wait, as the server is idle when they arrive.
The fourth arrival that joins at 9.0 minutes has to wait for 0.5 minute, while
the service to the third is completed. Similarly the waiting time and idle time
can be computed for further arrivals.
Total elapsed time = 29 minutes
Waiting time of arrival = 1 minute
Percentage of waiting time = ( 1 x 100) / 29 = 3.4
Idle time for server = 14.5 minutes
Percentage of idle time = ( 14.5 x 100) / 29 =50%
Sikkim Manipal University
Operations Research
Unit 12
Operations Research
Unit 12
Operations Research
Unit 12
11. Large number of trails require a great deal of computational effort and
the use of computer becomes optional. (True/False)
12. A practical indicator of when to stop simulation trials is when the results
that violently fluctuate initially tend to stabilise if the simulation is
continued. (True/False)
Solved problem 2:
Every morning a bread vendor buys loaves of bread at Re. 0.45 each by
placing his order one day in advance (at the time of receiving his previous
order) and sells them at Re. 0.70 each. Unsold bread can be sold the next
day at Re. 0.20 per loaf and thereafter should be treated as of no value. The
pattern of demand for bread is depicted in table 12.11:
Table 12.11: Pattern of Demand for Bread
Fresh Bread
Daily Sales
Probability
of demand
Daily
Sales
Probability of
demand
50
0.01
0.10
51
0.03
0.20
52
0.04
0.08
53
0.07
0.02
54
0.09
55
0.11
56
0.15
57
0.21
58
0.18
59
0.09
60
0.02
The vendor adopts the following order rule, if there is no stock with him at
end of the previous day. He orders 60 units. Otherwise he orders 50 or 55
whichever is nearest the actual fresh bread sale on the previous day.
Starting with zero stock and a pending order for 55 loaves simulate for 10
days and calculate the vendors profits.
Operations Research
Unit 12
Solution:
Table 12.12 depicts the allocation of random numbers.
Table 12.12: Allocation of Random Numbers
Fresh Bread
Daily
Prob
of
dema
nd
Cum Prob
of
demand
Rand
No.
Dail
y
Sale
Prob of
deman
d
Cum
Prob of
demand
allocated
50
0.01
0.01
00
0.70
0.70
51
0.03
0.04
01 03
0.20
0.90
52
0.04
0.08
04 07
0.08
0.98
53
0.07
0.15
08 14
0.02
1.00
54
0.09
0.24
15 23
55
0.11
0.35
24 34
56
0.15
0.50
35 49
57
0.21
0.71
50 70
58
0.18
0.89
71 88
59
0.09
0.98
89 97
60
0.02
1.00
98 - 99
Sale
Rand
No.
Alloca
-ted
00 to
69
70 to
89
90 to
97
98 to
99
You can now construct a table through simulation to see how the stocks and
sales fluctuate. Table 12.13 depicts the results of simulation.
Operations Research
Unit 12
55
60
60
50
55
60
60
55
55
60
72
06
12
14
79
70
85
71
21
98
Total
570
Sale
58
52
53
50
58
55
58
57
58
55
58
54
60
549
Order
for
next
day
Opening
stock
1
2
3
4
5
6
7
8
9
10
Day
Receipt
of the
start day
Random
No
FRESH BREAD
0
8
7
0
0
3
2
0
1
0
60
60
50
55
60
60
55
55
60
55
0
0
8
7
0
0
3
2
0
1
Random
No.
86
54
88
58
48
21
Sale
0
0
1
0
0
0
1
0
0
0
Operations Research
Unit 12
No. of failures
Age at failure
Inlet side
Outlet side
(hours)
Bearing (A)
Bearing (B)
150
300
16
450
18
600
23
12
750
14
16
900
10
18
1,050
33
1,200
1,350
The cost of bearing is Rs. 300 each for A and Rs. 500 each for B. The
cost of downtime of equipment is Rs. 700 per hour and it takes 2 hours to
replace one bearing either at inlet or outlet side and 3 hours to replace both
the bearings. The three maintenance policies to be evaluated are:
1. Replace a bearing only when it fails.
2. Replace both the bearings if one fails.
3. Replace the bearing which fails plus the other one if it has been in use
for more than its estimated average services life 600 hours for bearing
A and 860 hours for bearing B
Find the best alternative through simulation.
Solution:
It is assumed that the failure of a bearing is independent of the maintenance
policy followed. Random numbers are allocated for different failure times:
(Table 12.15 depicts the random numbers)
Operations Research
Unit 12
Bearing B
Cumulative
Random Nos.
Probability
Cumulative
Probability
Random
Nos.
150
0.07
00 to 96
0.00
300
0.23
07 to 22
0.03
00 to 02
450
0.41
23 to 40
0.12
03 to 11
600
0.64
41 to 63
0.24
12 to 23
750
0.78
64 to 77
0.40
24 to39
900
0.88
78 to87
058
40 to57
1050
0.95
88 to 94
0.91
58 to 90
1200
1.00
95 to 99
0.98
91 to 97
1350
1.00
98 and99
We can now select random numbers from the tables and generate a set of
12 bearings for each of the bearings. Table 12.16 depicts the set of 12
bearing for each of the bearings generated.
Table 12.16: Set of 12 Bearings for Each of the Bearings Generated
1
2
3
4
5
6
7
8
9
10
11
12
10
22
24
42
37
77
99
96
89
85
28
63
300
300
450
600
450
750
1200
1200
1050
900
450
600
300
600
1050
1650
2100
2850
4050
5250
6300
7200
7650
8250
99
96
18
36
50
79
80
96
31
07
62
77
1350
1200
600
750
900
1050
1050
1200
750
450
1050
1050
1350
2550
3150
3900
4800
5850
6900
8100
8850
9300
10,350
11,400
Let us compare the costs of three policies for the first 7200 hours.
Policy 1: Replace a bearing only when it fails.
A requires replacement 10 times, and B 7 times during this period as seen
from the lives of successive bearings.
Sikkim Manipal University
Operations Research
Unit 12
Bearing fails
first
Life (hrs) of
newly fitted
bearing A (From
table 12.16)
Life (hrs) of
newly fitted
bearing B (From
table 12.16)
300
300
1200
600
450
600
1050
600
750
1650
450
900
2100
750
1050
2850
1200
1050
3900
1200
1200
5100
A&B
1050
750
5850
900
450
6300
450
1050
6750
600
1050
7350
Operations Research
Unit 12
Start
of
the
period
Life bearing
in service
(hrs)
Be
aring
wh
ich
fail
s
Time
elapsed
(hrs)
Age
surviving
bearing
Did the
surviving
bearing
(A/B)
Complete
its est.
avg life?
If No
in col
7
balance life
of
surviving
Replace
ment
policy
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
300
1350
300
300
B - No
1050
Replace
A
300
300
1050
600
600
B - No
750
Replace
A
600
450
750
1050
1050
B - Yes
--
Replace
A&B
1050
600
1200
1650
600
B - No
600
Replace
A
1650
450
600
2100
1050
B - Yes
--
Replace
A&B
2100
750
600
2700
600
B - Yes
--
Replace
A&B
2700
1200
750
3450
750
B - Yes
--
Replace
A&B
3450
1200
900
4350
900
B - Yes
--
Replace
A&B
4350
1050
1050
A&
B
5400
--
--
Replace
A&B
5400
900
1050
6300
900
B - Yes
--
Replace
A&B
6300
450
1200
6750
450
B - No
750
Replace
A&B
6750
600
750
7350
1050
B - Yes
--
Replace
A&B
This will be the life (from Table 12.18) if newly fitted. For survivors from
previous replacement, this will be the balance life (col. 8).
Sikkim Manipal University
Operations Research
Unit 12
Solved problem 4:
A factory produces 150 scooters. But the production rate varies with the
distribution depicted in table 12.19.
Table 12.19: Production Rate and Probability
Production
Rate
147
148
149
150
151
152
153
Probability
0.05
0.10
0.15
0.20
0.30
0.15
0.05
At present the track will hold 150 scooters. Using the following random
numbers determine the average number of scooters waiting for shipment in
the factory and average number of empty space in the truck. Random
Numbers 82, 54, 50, 96, 85, 34, 30, 02, 64, 47.
Solution:
Table 12.20 depicts the production rate and probability.
Table 12.20: Production Rate and Probability
Production
rate
Probability
Cumulative
Probability
Random No.
Assigned
147
0.05
0.05
00 04
148
0.10
0.15
05 - 14
149
0.15
0.30
15 - 29
150
0.20
0.50
30 - 49
151
0.30
0.80
50 - 79
152
0.15
0.95
80 - 94
153
0.05
1.00
95 - 99
Random
No.
Simulated
Production
Rate
Scooter
Waiting in
the factory
1
2
3
4
5
6
82
54
50
96
85
34
152
150
150
153
152
150
2
2
2
5
7
7
Number of
example
spaces in
the truck
Page No. 246
Operations Research
7
8
9
10
Unit 12
30
02
64
47
Total
150
147
151
150
7
4
5
7
3
3
Probability
Filling
Check up
Crowning
Cleaning
Extraction
0.10
0.30
0.15
0.30
0.15
Solution:
Table 12.23 depicts the assigning of random number.
Table 12.23: Assigning Random Number
Illness
Probability
Cum. Prob.
Filling
Check up
Crowning
Cleaning
Extraction
0.10
0.30
0.15
0.30
0.15
0.10
0.40
0.55
0.85
1.00
Random No.
Assigned
00 -09
10 -39
40 -54
55 -84
85 -99
Operations Research
Unit 12
Starts
Finishes
Patients
Waiting
Time
Treatment
Doctor
Idle time
Patients
Arrival
Time
Nature of
illness
Time
Taken
Trial
No
Random
No.
56
Cleaning
15
8.00
8.00
8.15
40
Crowning
40
8.30
8.30
9.10
15
26
Check up
15
9.00
9.10
9.25
10
66
Cleaning
15
9.30
9.30
9.45
87
Extraction
30
10.00
10.00
10.30
15
48
Crowing
40
10.30
10.30
11.10
17
Check up
15
11.00
11.10
11.25
10
22
Check up
15
11.30
11.30
11.45
04
Filling
50
12.00
12.00
12.50
15
10
15
Check up
15
12.30
12.50
13.05
20
45
40
Total
12.7 Summary
Let us recapitulate the important concepts discussed in this unit:
Using simulation, an analyst can introduce the constants and variables
related to the problem, set up the possible courses of action and
establish criteria which act as measures of effectiveness.
The methodology developed for simulation process consists of seven
steps.
In any simulation problem, the variables to be studied will be given with
associated probabilities.
The mid square method is used for generating pseudorandom numbers.
One of the disadvantages of the mid square method is that the
Operations Research
Unit 12
12.8 Glossary
Sample size: Number of trials in a simulation
GPSS (General Purpose Simulation System): a discrete event simulation
programming system developed in 1950s for application in queuing,
manufacturing, transportation and defense systems
SIMSCRIPT: a discrete event simulation programming system developed in
1962 by Harry Markowitz .It can be programmed with simple English like
statements to generate the simulation model.
Operations Research
Unit 12
Random Numbers (10, 92); (25, 83); (36, 76); (44, 15); (57, 25); (62, 67);
(04, 99); (72, 53); (81, 35); (94, 07)
12.10 Answers
Self Assessment Questions
1. True
2. False
3. False
4. Management laboratory
5. Equally likely
6. 1
7. True
8. True
9. False
10. True
11. False
12. True
Terminal Questions
1. The methodology developed for simulation process consists of seven
steps. For more details refer to section 12.3
2. Concepts of simulation: simulation of complex problems, sampling
allotting random numbers to events and theoretical probability. For more
details refer to section 12.4
3. In any simulation problem, the variables to be studied will be given with
associated probabilities. The initial conditions will also be specified. You
can choose random numbers from table. For more details refer to
section 12.5
4. To generate a list of number and allot it to events. For more details refer
to section 12.5.2
5. 57.7 minutes. For more details refer to section 12.6
Operations Research
Unit 12
Operations Research
Unit 13
Unit 13
Structure:
13.1 Introduction
Objectives
13.2 Monte-Carlo Simulation
Problems on Monte-Carlo method
13.3 Applications of Simulation
Simulation in networks
Simulation in job sequencing
13.4 Advantages of Simulation
13.5 Limitations of Simulation
13.6 Summary
13.7 Glossary
13.8 Terminal Questions
13.9 Answers
13.10 Case Study
13.1 Introduction
In the previous unit, you learnt the basic concepts and methodology of
simulation. You also learnt simulation procedure and sample size. MonteCarlo technique has become such an important part of simulation models
that the terms are often assumed to be synonymous. However, it is only a
special technique of simulation. The technique of Monte-Carlo involves the
selection of random observations within the simulation model.
The technique is restricted for application involving random numbers to
solve deterministic and stochastic problems. The principle of this technique
is replacement of actual statistical universe by another universe described
by some assumed probability distribution and then sampling from this
theoretical population by means of random numbers.
In fact, this process involves the generation of simulated statistics (random
variables) that can be explained in simple terms as choosing a random
number and substituting this value in it.
In this unit, you will learn the Monte-Carlo simulation method and the
application of simulation methods. You will also learn the advantages and
limitations of simulation.
Sikkim Manipal University
Operations Research
Unit 13
Objectives:
After studying this unit, you should be able to:
describe Monte-Carlo simulation method
analyse the applications of simulation
list the advantages and limitations of simulation
Operations Research
Unit 13
Operations Research
Unit 13
No
rain
1 cm
Rain
2 cm
Rain
3 cm
Rain
4 cm
Rain
5 cm
Rain
Probability
0.50
0.25
0.15
0.05
0.03
0.02
If it did not rain on the previous day, the rain distribution is as depicted in
table 13.2:
Table 13.2: Rain Distribution Table
Event
Probability
No rain
1 cm
Rain
2 cm
Rain
3 cm
Rain
0.75
0.15
0.06
0.04
Simulate the citys weather for 10 days and determine the total days without
rain as well as the total rainfall during the periodby simulation. Use the
following random numbers:
67
63
39
55
29
78
70
06
78
76
For simulation assume that for the first day of the simulation it had not
rained the day before.
Solution: We simulate the citys weather with and without rainfall in the
following steps:
Step 1: Previous day rain distribution is as depicted in table 13.3:
Table 13.3: Rain Distribution Table
Event
Probability
No rain
1 cm Rain
2 cm rain
3 cm Rain
4 cm Rain
5 cm Rain
0.50
0.25
0.15
0.05
0.03
0.02
Cumulative
probability
0.50
0.75
0.90
0.95
0.98
1.00
RN range
00-49
50-74
75-89
90-94
95-97
98-99
Page No. 255
Operations Research
Unit 13
Probability
Cumulative probability
RN range
No rain
0.75
0.75
00-74
1 cm Rain
0.15
0.15
75-89
2 cm rain
0.06
0.96
90-95
3 cm Rain
0.04
1.00
96-99
RN
Events
Cumulative rain
67
No rain
63
No rain
39
No rain
55
No rain
29
No rain
78
1 cm Rain
1 cm
70
1 cm Rain
2 cm
06
No rain
2 cm
78
1 cm Rain
3 cm
10
76
2 cm Rain
5 cm
During the simulated period it did not rain on 6 out of 10 days. The total
rainfall during the period is 5 cm.
Solved problem 2:
The number of customers at a restaurant each evening is distributed as
depicted in table 13.6:
Table 13.6: Distribution of Customers
Number of customers
Lots
Average
Very few
Probability
0.2
0.4
0.4
The chef refuses to work on an evening when there are very few customers
and walks out. He will not return to work until an evening when there are lots
Sikkim Manipal University
Operations Research
Unit 13
Very few
0.4
6,7,8,9
Operations Research
Unit 13
The following specimen run was started with Monday and with the chef
being in. Keeping the view in mind that we are trying to reproduce the long
term behaviour of the restaurant, the initial conditions of the simulation will
not affect the results we obtain.
Day
Day of the
week
1
2
3
4
5
6
7
8
9
10
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday
Monday
Tuesday
Wednesday
Average
Average
Lots
Very few
Very few
Very few
Average
Average
Lots
In
In
In
Out
In
Out
Out
Out
In
Number
of days
in to
date
1
2
3
3
4
4
4
4
5
Fraction
of days
in to
date
1.00
1.00
1.00
0.75
0.80
0.67
0.57
0.50
0.56
Operations Research
Unit 13
Operations Research
Unit 13
Solved problem 3:
A project schedule consists of five activities and the duration of these
activities is non-deterministic with the probability distribution as depicted in
table 13.9:
Table 13.9: Probability Distribution Table
Activity
Days
Probability
1-2
1
4
8
0.2
0.5
0.3
1-3
2
4
7
0.3
0.5
0.2
2-4
2
4
6
0.3
0.3
0.4
3-4
3
6
8
0.3
0.4
0.3
4-5
2
3
4
0.2
0.2
0.6
Simulate the duration of the project 10 times and estimate the chances of
various paths to be critical. What is the average duration of the project?
Solution:
From the PERT network, we observe that there are two paths from the start
to end of the project, namely 1245 and 1345. Using one-digit
random number corresponding to the probability distribution of various
activities, we obtain the simulated duration of each path and hence the
simulated duration of various activities and the simulated duration of the
project. Table 13.10 depicts the results of the simulation:
Operations Research
Unit 13
1-2
RN D
1-3
RN D
2.4
RN D
3-4
RN D
4-8
RN D
Project
Duration
1
2
3
8
5
5
8
4
4
9
5
6
2
4
4
8
8
7
6
6
6
3
1
4
6
3
6
6
7
7
4
4
4
18
14
14
4
5
6
7
8
9
10
7
0
8
4
5
2
6
8
1
8
4
4
4
4
5
6
3
7
8
5
1
4
4
4
4
7
4
2
1
8
0
4
2
9
3
2
6
2
4
2
6
4
9
1
4
4
6
0
7
8
3
6
0
6
3
8
5
8
3
3
6
3
0
4
4
3
3
4
3
2
16
11
13
11
17
13
12
Critical Paths
1245
1245
1245
1345
1345
1245
125
145
145
145
145
Immediate
predecessor
A
B
B,C
E,F
7
0.20
8
0.10
6
0.30
Operations Research
Unit 13
A) Draw the network diagram and identify the critical path using the
expected activity times.
B) Simulate the project using random numbers to determine the activity
times. Find the critical path and the project duration.
C) Repeat the simulation four more times. Is the same path critical in all the
simulations?
Solution:
a) using the given information, the resulting network is depicted in figure
13.2. The time for each of the activities is the expected time obtained by the
summation of the products of the activity time and the corresponding
probability. For example, for activity A,
The Expected Time = 3*0.20 + 4*0.60 + 5*0.20 = 4 days.
Network design
Observe that the critical path of the project is 123467, with an
expected duration of 20.2 days.
To use the simulation for obtaining the project length, we first determine the
random number intervals for each activity. Random number coding is shown
in the table. The five simulation runs, by taking random numbers, are as
depicted in table 13.12. For each run, the critical path and the project
durations are obtained and the project time is obtained as follows:
Total time = Larger of times for activities A and C, and B + Time for activity
D + Larger of times for activities E + Time for activity G
Operations Research
Unit 13
Time
Probability
Cumulative
Probability
Random
number
Interval
3
4
5
0.20
0.60
0.20
0.20
0.80
1.00
00-19
20-79
80-99
4
5
6
7
8
0.10
0.30
0.30
0.20
0.10
0.10
0.40
0.70
0.90
1.00
00.09
10-39
40-69
70-89
90-99
1
3
5
0.15
0.75
0.10
0.15
0.90
1.00
00-14
15-89
90-99
4
5
3
0.80
0.20
0.10
0.80
1.00
0.10
00-79
80-99
00-09
4
5
6
0.30
0.30
0.30
0.10
0.70
1.00
10-39
40-69
70-99
5
7
0.20
0.80
0.20
1.00
00-19
20-99
2
3
0.50
0.50
0.50
1.00
00-49
50-99
Simulation worksheet
Activity
Table 13.13 depicts the simulation worksheet.
Table 13.13: Simulation Worksheet
Simulation
1
2
3
4
5
A
RN D
68
99
57
57
77
4
5
4
4
4
B
RN D
13
93
33
12
37
5
8
5
5
5
C
RN D
09
18
49
31
34
1
3
3
3
3
D
RN D
20
20
65
96
11
4
4
4
5
4
E
RN
73
22
92
86
27
F
D
RN
6
4
6
6
4
07
07
98
92
10
G
D
RN
5
5
7
7
5
92
29
00
91
59
3
2
2
3
3
Operations Research
Unit 13
We get the information as depicted in figure 13.3 from the above table:
It may be observed that the same path is not critical in each of the runs.
13.3.2 Simulation in job sequencing
In this section, we illustrate the use of simulation in the sequencing
problems.
Example:
A job has to be processed over two machines, machine M1 and machine
M2 in that order. The distribution of inter-arrival time of the jobs at the first
machine is as depicted in table 13.14:
Table 13.14: Distribution of Arrival Time
Time(minutes)
Probability
0.2
0.2
0.2
0.4
Operations Research
Unit 13
Probability
Time(minutes)
Probability
0.1
0.2
0.2
0.3
0.3
0.4
0.3
0.1
0.1
On the basis of 10 simulation runs, find out the average queue length before
machine M1 and the average queue length before machine M2.
Solution:
The various steps for simulation are:
1. Simulate the inter-arrival time of the ith job on machine M1 with the help
of a random number (i= 1, 2.).
2. Arrival time of ith job.
= Arrival time of (i-1)th job (i=2,3, .) + Inter-arrival time of ith job.
Arrival time of the first job= Inter-arrival time of the first job
3. Simulate the processing time on machine M1 by random numbers.
4. Departure time on machine M1
= Arrival time for processing on machine M2
= Max (Arrival time of ith job, process completion time of (i-1)thjob
on machine M1) + processing time of ith job on machine M1
5. If arrival time of ith job < Process completion time of (i-1)th, ith arrival
waits.
If arrival time of ith job < Process completion time of (i-2)th, both ith and (i1)th arrivals wait.
6. Simulate the processing time of machine M2 with the help of random
numbers.
7. Process completion by machine M2 of ith arrival
= Max [Arrival time of ith job, process completion time of (i-1)th job]
+ Process time by machine M2.
Sikkim Manipal University
Operations Research
Unit 13
8. If arrival time of ith< process completion time of (i-1)th, the ith arrival waits.
If arrival time of ith< process completion time of (i-2)th , both ith and (i-1)th
arrivals wait, etc.
Using the procedure, the simulated processing data of 10 arrivals are
displayed in the simulation worksheet as depicted in figure 13.4:
On the basis of the simulation study, the average queue length before
machine M1 is 0.5 and the average length before machine M2 is 1.2.
Self Assessment Questions
3. The ___________ and the project duration can be found out with a
randomly selected activity times for each activity.
4. The range of application of simulation in _________ is extremely wide.
Operations Research
Unit 13
Operations Research
Unit 13
13.6 Summary
Let us recapitulate the important concepts discussed in this unit:
Operations Research
Unit 13
13.7 Glossary
Job sequencing: the arrangement of the tasks required to be carried out.
job shop: Organization of work in a manufacturing organization by job
functions.
Simulate the project and determine, using random numbers, the activity
times. Find the critical path and the project duration. Repeat five times.
Does the critical path change? State the estimated duration of the project in
each of the trials.
Activity
Time days
1
0.02
0.04
0.4
0.3
0.3
0.2
0.3
0.6
H
Sikkim Manipal University
0.7
0.4
0.6
0.8
0.3
0.1
0.2
E
G
0.5
0.2
0.2
0.3
0.4
0.2
0.4
Page No. 269
Operations Research
Unit 13
13.9 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
Terminal Questions
1. To find the critical path and project duration. For more details refer
13.3.1
2. Advantages of simulation: to studying complicated systems, investigate
the consequences for a system and assess the possible risks involved.
For more details refer 13.4
3. Limitations of simulation: does not generate the solution techniques,
simulation models are expensive and does not produce answers by
itself. For more details refer 13.5
4. The Monte-Carlo method is a simulation technique in which statistical
distribution functions are created by using a series of random numbers.
This approach has the ability to develop many months or years of data
in a matter of few minutes on a digital computer. For more details
refer 13.2
5. Using random numbers from the first six columns for six trials:
Trial
Critical paths
Duration
1
BDFGH
16
2
BDFGH
19
3
BDFGH
18
4
BDFGH
17
5
ACFGH
BDFGH
16
6
ACFGH and
BDFGH
18
Sikkim Manipal University
Operations Research
Unit 13
Operations Research
Unit 13
Operations Research
Unit 13
waits are not excessive. To avoid excessive waits for either kind of casting,
all the castings are processed as soon as possible on a first come, firstserve
basis.
Discussion Question:
Managements objective is to minimise the expected total cost per hour. Use
simulation to evaluate and compare all the alternatives, including the status
quo and the various combinations of proposals. Then make your
recommendation to the management.
References:
Kanti Swarup & Gupta P. K., & Hira D. S., & Manmohan (2004).
Operation Research. Sultan Chand and Sons.
Operations Research
Unit 14
Unit 14
Structure:
14.1 Introduction
Objectives
14.2 Basic Difference between PERT and CPM
PERT
CPM
Project scheduling by PERT-CPM
14.3 PERT/CPM Network Components and Precedence Relationship
Critical path calculations
Determination of the critical path
Determination of floats
14.4 Project Management PERT
14.5 Summary
14.6 Glossary
14.7 Terminal Questions
14.8 Answers
14.9 Case Study
14.1 Introduction
In the previous unit, simulation Monte-Carlo method, we studied about the
Monte-Carlo simulation. We also learnt about the various problems of this
method, the applications of the simulation, the advantages and the
limitations of the simulation. In this unit, project scheduling and PERT-CPM
in operations research management, we will study about the basic
difference between PERT and CPM. We will also learn about PERT and
CPM network components, determination of floats and project management.
Project management has evolved as a new field with the development of
two analytic techniques for planning, scheduling and controlling projects.
These are the project evaluation and review technique (PERT) and the
critical path method (CPM). PERT and CPM are basically time-oriented
methods in the sense that they both lead to the determination of a time
schedule.
Projects may be defined as a collection of interrelated activities (or tasks)
which must be completed in a specified time according to a specified
Sikkim Manipal University
Operations Research
Unit 14
Operations Research
Unit 14
14.2.2 CPM
Operations Research
Unit 14
Operations Research
Unit 14
Operations Research
Unit 14
Operations Research
Unit 14
Solution
Figure 14.5 depicts an analysis network.
Operations Research
Unit 14
Operations Research
Unit 14
We can now identify the critical path activities using the results of the
forward and backward passes. An activity (i, j) lies on the critical path if it
satisfies the following conditions:
ESi = LCi
ESj = LCj
ESj-ESi = LCj-LCi = Dij
These conditions actually indicate that there is no float or slack time
between the earliest start and the latest start of the activity. Thus, the
activity must be critical.
In the arrow diagram these are characterised by same numbers within
rectangles and triangles at each of the head and tail events. The difference
between the numbers in rectangles or triangles at the head event, and the
number within rectangles or triangles at the tail event is equal to the duration
of the activity. Thus, we will get a critical path, which is a chain of connected
activities, spanning the network from start to end.
Solved problem 2
Consider a network which stands from node 1 and terminates at node 6,
the time required to perform each activity is indicated on the arrows.
Figure 14.6 depicts an analysis network.
Solution
Let us start with the forward pass with ESi = 0.
Since there is only one incoming activity (1, 2) to event 2 with D12 = 3.
ES2 = ES1+ DS2 = 0+3=3.
Operations Research
Unit 14
Let us consider the end 3, since there is only one incoming activity (2, 3)
to event 3, with D23 = 3.
ES3 = ES2+ D23 = 3+3 = 6.
To obtain ES4, since there are two activities A (3, 4) and (2, 4) to the
event 4 with D24 = 2 and D34 = 0.
ES4= maxi=2, 3 {ESi + De4}
= max {ES2 +D24, ES3 + D34}
= max {3+2, 6+0} = 6
Similarly, ES5 = 13 and ES6 = 19.
This completes the forward pass.
In the backward pass we have
LC6= 19 = ES6
LC5 = 19-6 = 13
LC4= min J = 5, 6 {LCj D4j} = 6
LC3 = 6, LC2 = 3 and LC1 = 0
Therefore, activities (1, 2), (2, 3) (3, 4) (4, 5) (5, 6) are critical and (2, 4)
(4, 6), (3, 6), are non-critical.
Thus, the activities (1, 2), (2, 3), (3, 4), (4, 5) and (5, 6) define the critical
path which is the shortest possible time to complete the project i.e., 19
days.
14.3.3 Determination of floats
Once the determination of the critical path is done, we have to compute the
floats for the non-critical activities. For the critical activities this float is zero.
Before showing how floats are determined, it is necessary to define two new
times that are associated with each activity. These are as follows:
Latest start (LS) time and
Earliest completion (EC) time
We can define activity (i, j) for these two types of time by:
LSij= LCj Dij
ECij = ESi + Dij
There are two important types of floats namely:
Total float (TF)
Free float (FF)
Sikkim Manipal University
Operations Research
Unit 14
The total float TFij for activity (i, j) is the difference between the maximum
time available to perform the activity (= LCj ESi) and its duration (= Dij)
TFij = LCj ESi Dij = LCj ECij = LSij ESi
The free float is defined by assuming that all the activities start as early as
possible. In this case FFij for activity (i, j) is the excess of available time
(= ESi ESi) over its deviation (= Dij); i.e.
FFij = ESi ESi - Dij
Note: For critical activities float is zero. Therefore, the free float must be
zero when the total float is zero. However, the converse is not true, that is, a
non-critical activity may have zero free floats.
Let us consider the example taken before the critical path calculations. The
floats for the non-critical activities can be summarised as depicted in the
table 14.1.
Activity
(i j)
Free
Float
FFij
(1, 2)
0*
(2, 3)
0*
(2, 4)
(3, 4)
0*
(3, 5)
10
13
(3, 6)
17
19
11
11
(4, 5)
13
13
0*
(4, 6)
11
14
19
(5, 6)
13
19
13
19
0*
Operations Research
Unit 14
Solved problem 3
A project consists of a series of tasks A, B, C, D, E, F, G, H, I with the
following relationships:
W < X,Y means X and Y cannot start until W is completed
X,Y < W means W cannot start until both X and Y are completed
With this notation construct the network diagram having the following
constraints A < D, E; B, D < F; C < G, B < H; F, G < I.
Also find the minimum time of completion of the project, the critical path,
and the total floats of each task, when the time (in days) of completion of
each task is as follows:
Task: A
B
C
D
E
F
G
H
I
Time: 23 8
20 16 24 18 19 4
10
Solution
Figure 14.7 depicts an analysis network.
ES1 = 0, ES2 = 20, ES3 = 23, ES4 = 59, ES5 = 39, ES6 = 57, ES7 = 67
Table 14.2 depicts an activity table.
Table 14.2 Activity Table
Activity Duration
(i, j)
Dij
Start
ESe
Finish
Eeij
Start
Lj-Dij
Finish
Lj
Table
Float
TFij
Earliest
Latest
Free
Float
FFij
(1, 2)
20
20
18
38
18
(1, 3)
23
23
23
0*
(1, 4)
31
39
31
31
(2, 5)
19
20
39
38
57
18
(3, 4)
16
23
39
23
39
0*
Operations Research
Unit 14
(3, 7)
24
23
47
43
67
20
20
(4, 5)
39
39
57
57
10
(4, 6)
18
39
57
39
57
0*
(5, 6)
39
39
57
57
18
18
(5, 7)
39
43
63
67
24
24
(6, 7)
10
37
67
57
67
0*
Critical path is 1 3 4 6 7.
Self Assessment Questions
6. Events do not consume ________ and _________.
7. Arrows head number is _________ than its tail number.
8. Dummy activity is introduced in a network to keep proper _________
relationship.
9. Critical path calculation includes both _________ and _________.
Operations Research
Unit 14
Therefore,
ab
2m
a b 4m
2
D
3
6
We can use this estimate to study the single estimate D in the critical path
calculation. These formulae are derived from beta distribution.
The variance of each activity denoted by V is defined by,
ba
Variance V =
The earliest expected times for the node i is denoted by E (i). For each
node i, E(i) is obtained by taking the sum of the expected times of all
activities leading to the node i, when more than one activity leads to a node
i, then the greatest of all E(i) is chosen. Let i be the earliest occurrence
time of the event i, we can consider i as a random variable. Assuming that
all activities of the network are statistically independent, we can calculate
the mean and the variance of i as follows:
E {i } = ESi and Var{i } =
Vk
k
Operations Research
Unit 14
E( )
STi E( i )
i
Pr (i STi) = Pr i
V( i )
V( i )
= Pr (Z Ki)
Where, Z N (01) and Ki =
STi E( i )
V ( i )
Optimistic time
18
26
16
15
Pessimistic time
10
22
40
20
25
12
12
20
33
18
20
10
Operations Research
Unit 14
Solution
Using the formula we can calculate expected activity times and variance
as depicted in table 14.4.
D
b a 2
(a b 4m) V
6
6
1
A
5
18
26
16
15
6
7
7
3
Forward pass
E1 = 0
E2 = 7.8
E7 = 42.8
B
10
22
40
20
25
12
12
9
5
E3 = 20
m
8
20
33
18
20
9
10
8
4
7-8
20-00
33-0
18-0
20-0
9-0
9-8
8-0
4-0
E4 = 33
v
0.696
0.444
5.429
0.443
2.780
1.000
0.694
0.111
0.111
E5 = 25.8
Backward pass
L7 = 42.8 L6 = 38.8 L5 = 34.8 L4 = 33.0 L3 = 29.8
L1 = 0.
The E-values and L-values are depicted in figure 14.9.
E6 = 29
L2 = 16.8
The critical path is shown by a thick line in the figure. The critical
path is 1-4-7 and the earliest completion time for the project is 42.8
weeks.
The last event 7 will occur only after 42.8 weeks. For this we require
only the duration of critical activities. This will help us in calculating
the standard duration of the last event.
Operations Research
Unit 14
14.5 Summary
Critical path computations are quite simple, yet they provide valuable
information that simplifies the scheduling of complex projects.
The result is that PERT-CPM techniques enjoy tremendous popularity
among practitioners in the field.
The usefulness of the techniques is further enhanced by the availability
of specialised computer systems for executing, analysing and controlling
network projects.
14.6 Glossary
Uncertainty: volatility.
Task: activity that has to be completed within the specified time.
Sikkim Manipal University
Operations Research
Unit 14
Latest start (LS) time: the latest time that an activity can begin without
delaying the project completion time.
Preceding activity
A
A
A
C
D
B, D, E
H
I, G, F
Duration in weeks
6
3
5
4
3
3
5
5
2
3
Preceding
activity
A
B
C
D
E
F
G
H
None
None
A
A
A
B, C
D
E, F, G
Most
optimistic
time (a)
2
10
8
10
7
9
3
5
Most likely
time (m)
4
12
9
15
7.5
9
3.5
5
Most
pessimistic
time (b)
12
26
10
20
11
9
7
5
Page No. 291
Operations Research
Unit 14
14.8 Answer
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
True
True
True
False
True
Time, resource
Greater than
Precedence
Forward pass & backward pass
False
True
True
True
True
Terminal Questions
1. There are no essential differences between PERT and CPM as both of
them share in common the determination of a critical path. For more
details refer to section 14.2
2. Probability considerations are incorporated in PERT by assuming that
the time estimate for each activity is based on three time estimates of
optimistic time, pessimistic time and most likely time. For more details
refer to section 14.4
3. b) 20 weeks c) A D H I - J . For more details refer section 14.3
4. b) A D G H c) 0.6591 d) 34.7 weeks. For more details refer
section 14.4
Sikkim Manipal University
Operations Research
Unit 14
Expected
Activities
12
8
6
13
4
101
6
101
8
8
4
Precedent
Minimum
Crash
Time(weeks)
(rs/week)
10
24,000
A
18
-11
A
11
3,000
A
13
-11
B,C
4
-11
C
8
30,000
C
4
1500
E,F
6
75,000
D,G
8
-11
H,I
8
-11
J
4
-11
Page No. 293
Operations Research
Unit 14
Kanti Swarup & Gupta P. K., & Hira D. S., & Manmohan (2004).
Operation Research. Sultan Chand and Sons.
Operations Research
Unit 15
Unit 15
Game Theory
Structure:
15.1 Introduction
Objectives
15.2 Competitive Situations
Marketing different brands of a commodity
Campaigning for elections
Fighting military battles
15.3 Characteristics of Competitive Games
N-person game
Zero-sum game
Two-person zero-sum game (Rectangular game)
Strategy
Pure strategy
Mixed strategy
15.4 Maximin Minimax Principle
Saddle point
Solution to a game with a saddle point
15.5 Dominance
Solving games using dominance
15.6 Summary
15.7 Glossary
15.8 Terminal Questions
15.9 Answers
15.10 Case Study
15.1 Introduction
In the previous unit, we learnt about project management and basic
difference between PERT and CPM. We also learnt the PERT/CPM network
components and precedence relationship.John Von Newman developed the
Game Theory in the year 1928. The theory gained importance after 1944,
when John Von Newman along with Morgenstern published their work
onTheory of games and economic behaviour. This highly resourceful field
of study is developing fast.
Operations Research
Unit 15
In this unit, you will learn about game theory, competitive situations and
characteristics of competitive games. You will also learn about Maximin
Minimax principle and dominance.
Objectives:
After studying this unit, you should be able to:
analyse competitive situations
apply game theory to such situations
evaluate a solution using saddle point and dominance principle
define zero-sum game
Operations Research
Unit 15
Operations Research
Unit 15
Operations Research
Unit 15
A1 (one finger)
A2 (two fingers)
-10
5
5
-10
Operations Research
Unit 15
max
i
min
j ( a ij )
Operations Research
Unit 15
max
i
( a ij ) .
min
j
max
i ( a ij )
If v ( say ) , the maximin and the minimax are equal and the game is
said to have a saddle point. If , then the game does not have a saddle
point.
Note: cannot be greater than .
15.4.1 Saddle point
In a two-person zero-sum game, if the maximin and the minimax are equal,
the game has a saddle point.
Saddle point is the position where the maximin (maximum of the row
minimums) and minimax (minimum of the column maximums) coincide.
If the maximin occurs in the r th row and if the minimax occurs in the
Operations Research
Unit 15
Note 2: If a game does not have a saddle point, the solution offers a mixed
strategy.
Note 3: A game is said to be fair if its value is zero.
15.4.2 Solution to a game with a saddle point
Consider a two-person zero-sum game with players A and B. Let
A 1 , A 2 ,... A m be the courses of action for player A. Let B 1 ,B 2 ,.....B n be the
courses of action for player B.
The saddle point of the game is as follows:
1. The minimum pay-off in each row of the pay-off matrix is encircled as
shown in solved problem 2.
2. The maximum pay-off in each column is written within a box as shown in
solved problem 2.
3. If any pay-off is circled as well as boxed, that pay-off is the value of the
game. The corresponding position is the saddle point.
Let (r, s) be the saddle point. Then, the suggested pure strategy for
player A is Ar . The suggested pure strategy for player B is B s . The value of
the game is a rs .
Note: However, if none of the pay-offs is circled or boxed, the game does
not have a saddle point. Hence, the suggested solution for the players is
mixed strategy.
Solved problem 1:
Two persons A and B, without showing each other, place a coin each on the
table. If the coins happen to be of the same denomination, player A will take
both of them. If they happen to be of different denominations, player B will
take both of them. Player A has a few one-rupee coins and two-rupee coins.
Player B has a few one-rupee, two-rupee and five-rupee coins.
i) Write down the pay-off matrix of A. Does the game have a saddle
point? If so, write down the solution.
ii) What happens to the game if both the players play only with one-rupee
and two-rupee coins?
Operations Research
Unit 15
Solution:
i) The pay-off matrix of A is as follows.
Here, the game does not have a saddle point. Therefore, the suggested
solution is adoption of mixed strategy for the players.
Operations Research
Unit 15
Solved problem 2:
Verify whether the 2-finger morra game explained earlier has a saddle
point. If so, write down the solution for the game.
Solution:
The pay-off matrix of player A is as follows:
Solved problem 3:
A labour union of a firm is negotiating a new five-year settlement
regarding payments with the management. The union has the following
options: A1 : Aggressive bargaining, A 2 : Bargaining with reasoning and
A 3 : Conciliatory approach. The likely mode of responses from the
management are: B 1 : Aggressive bargaining, B2 : Bargaining with
reasoning, B 3 : Legalistic approach and B 4 : Conciliatory approach.
The gains to the union in each case are as follows:
Operations Research
Unit 15
What strategy would you suggest for the two sides? What is the value of
the game?
Solution:
c)
Operations Research
Unit 15
Solved problem 4:
Solve the game.
Operations Research
Unit 15
Solved problem 5:
A two-person zero-sum game, has the following pay-off matrix. Solve the
game.
Operations Research
Unit 15
15.5 Dominance
In a rectangular game, the pay-off matrix of player A is pay-off in one
specific row ( r th row ) exceeding the corresponding pay-off in another
specific row ( s th row ) . This means thatwhatever course of action is
adopted by player B, for A, the course of action A r yields greater gains than
the course of action A s . Therefore, A r
th
th
equal
to)
the
corresponding
pay-off
in
the q th column ,
B p dominates B q .
At times, a convex combination of two or more courses of action may
dominate another course of action. Whenever a course of action (say
A s or B q ) is dominated by others, then that course of action ( A s or B q ) can
be deleted from the pay-off matrix. Such a deletion will not affect the choice
of the solution, but it reduces the order of the pay-off matrix. Successive
reduction of the order using dominance property helps in solving games.
15.5.1 Solving games using dominance
Consider a two-person zero-sum game
with
players A and B.
Let A1 , A 2 ,...... A m be the courses of action for player A. Let B1 , B 2 , ... B n be
the courses of action for player B.
Sikkim Manipal University
Operations Research
Unit 15
Suppose the game has a saddle point. Use the dominance property in
sequence to delete the courses of action of A as well as B till the pair
comprising the saddle point remains alone. The procedure to arrive at the
saddle point is as follows:
a) In the pay-off matrix, if each pay-off in r
th
th
As is deleted.
b) In the pay-off matrix, if each pay-off in p th column is less than (or equal
to) the corresponding pay-off in the q th column , B p dominates B q ,
hence B q is deleted.
c) Repeat the above steps in succession to achieve the saddle point.
Note: Sometimes, a convex combination of two or more courses of action
may dominate another course of action.
Solved problem 6:
Solve the following game using dominance property.
Solution: In the pay-off matrix, each pay-off in the first row exceeds the
corresponding pay-off in the third row. Therefore, A1 dominates A 3 . So,
A 3 is deleted. Hence the reduced matrix is as follows:
Here, each pay-off in the third column is less than the corresponding
pay-off in the first column. Therefore, B 3 dominates B 1 . Similarly,
Sikkim Manipal University
Operations Research
Unit 15
Here, since 12>8, A1 dominates A2 . And so, finally the matrix reduces
to
Thus, (1, 3) is the saddle point and the solution to the game is as follows:
a) Strategy for A is A1 .
b) Strategy for B is B 3 .
c) Value of the game is v = 12
Solved problem 7:
Solve the following zero-sum game and find its value.
Solution: In the pay-off matrix, each pay-off in the second column is less
than (or equal to) the corresponding pay-off in the third column. And so,
the course of action Q dominates R. Similarly, Q dominates S. After
deleting R and S, the reduced matrix is as follows.
Here, pay-off in the second row is greater than (or equal to) the
corresponding pay-offs in the first, third as well as fourth rows. Therefore,
Sikkim Manipal University
Operations Research
Unit 15
Solved problem 8:
In a two-person zero-sum game, the pay-off matrix of A is
Operations Research
Unit 15
15.6 Summary
Let us recapitulate the important concepts discussed in this unit:
Competitive situations arise when two or more parties with conflicting
interests operate.
Competitive situations arise when two or more parties with conflicting
interests operate.
In a game, the strategy of a player is predetermined. The player uses
this strategy to select a course of action during the game.The strategy of
a player may be pureor mixed.
The maximum of the row minimum in the pay-off matrix is called
maximin. The minimum of the column maximums in the pay-off matrix is
called minimax.
Saddle point is the position where the maximin (maximum of the row
minimums) and minimax (minimum of the column maximums) coincide.
Whenever a course of action is dominated by others, then that course of
action can be deleted from the pay-off matrix. The deletion will not affect
the choice of the solution, but it reduces the order of the pay-off matrix.
Successive reduction of the order using dominance property helps in
solving games.
15.7 Glossary
Matrix: rectangular arrangement of numbers in rows and columns
Transpose of a matrix:the matrix that results from interchanging rows into
and columns into rows
Strategy: course of action during the game for achieving gains
Sikkim Manipal University
Operations Research
Unit 15
15.9 Answers
Self Assessment Questions
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Operations Research
Unit 15
Terminal Questions
1. Value of the game is 1. As strategy is (0, 1, 0, 0), while Bs strategy is
(0, 0, 1, 0, 0).For more details refer section 15.3.3.
2. Value is 5. As strategy is (1, 0), while Bs strategy is (1, 0).For more
details, refer section 15.3.3.
3. Number of players is finite, each player has finite moves, player can
adapt any one course of action. For more details refer to section 15.3 for
answer.
4. The maximin is:
max
i
min
j ( a ij )
min
j
max
i ( a ij )
Operations Research
Unit 15
Entry
Entry
1
Butterfly stroke
1:01.6
Ram
Shyam
59.1
57.5
58.4
1:03.2
59.8
Backstroke
1:06.8
1:05.6
1:03.3
1:02.6
1:04.9
1:04.1
Breaststroke
1:13.9
1:12.5
1:04.7
1:06.1
1:15.3
1:11.8
The points awarded are 5 points for first place, 3 points for second place, 1
point for third place, and none for lower places. Both coaches believe that all
swimmers will essentially equal their best times in this meet. Thus Ram and
Shyam each will definitely be entered in two of the three events.
Discussion Questions
a) The coaches must submit all their entries before the meet without
knowing the entries for the other team, and no changes are permitted
later. The outcome of the meet is very uncertain. Formulate this as a
two-person, zero-sum game.
b) Eliminate dominated strategies, find the strategies for both the teams
according to minimax criterion.
References:
E-References:
http://www.springerlink.com/content/g7ph5p3k4128t608/?MUD=MP,
retrieved on Dec 20, 2011.