MBAD 2123 SLM (PRESS) Quatitive Techniques
Uploaded by
aishwarkMBAD 2123 SLM (PRESS) Quatitive Techniques
Uploaded by
aishwarkDIRECTORATE OF DISTANCE EDUCATION
M.B.A
MASTER OF BUSINESS
ADMINISTRATION
MBAD 2123
1923
QUANTITATIVE TECHNIQUES
Second Semester
Semester – II
SRM INSTITUTE OF SCIENCE AND TECHNOLOGY,
DIRECTORATE OF DISTANCE EDUCATION
Potheri, Chengalpattu District 603203,
Tamil Nadu,INDIA
Phone: 044 – 27417040 / 41
Website: www.srmist.edu.in / Email:office.dde@srmist.edu.in
EXPERT COMMITTEE
NAME DESIGNATION ORGANISATION
S.No.
1 Dr.R.Rajagopal Director DDE - SRMIST
2 Dr.V.M.Ponniah Dean - Management FOM - SRMIST
FOM – SRMIST, KTR
3 Dr.G.Venugopalan Academic Coordinator
FOM – SRMIST, KTR
4 Dr.T.Ramachandran Head of II year MBA
FOM – SRMIST, KTR
5 Dr.V.M.Shenbagaraman Head of I year MBA
Course Writer(s) Dr. L. AKILA
Information contained in this book has been obtained by its Author(s) from sources believed to be
reliable and are correct to the best of their knowledge. However Publishers and the Author(s) shall in
no event be liable for any errors, omissions or damages arising out of this information and specifically
disclaim any implied warranties or merchantability or fitness for any particular.
DIRECTORATE OF DISTANCE EDUCATION
SRM Institute of Science and Technology,
Potheri, Kancheepuram District 603203, Tamil Nadu,INDIA
Phone: 044 – 27417040 / 41
Website: www.srmist.edu.in / Email:office.dde@srmist.edu.in
MBAD 2123 Quantitative Techniques - Syllabus
CHAPTER 1
Introduction to Linear Programming(LP) - LP Formulations - Graphical Solutions - Simplex
Method to solve LP problems - Duality Concepts - Mathematical Formulation of Integer
Programming (IP) - Basic problem in Gomory’s cutting planeMethod - Dynamic
Programming: Algorithm - Shortest Route Problem
CHAPTER 2
Transportation Problem - Assignment Problem – Trans-shipment problem - Travelling
salesman problem
CHAPTER 3
Network Models - Shortest Path Problem - Maximum Flow Problem –Minimum Spanning
Tree - Sequencing: M Job 2 Machine and 3 Machine Problems -Decision Theory: Decision
making under Certainty and Uncertainty
CHAPTER 4
Queuing Theory: M.M/I: FCFS/N and FCFS/ Models-∞ - Replacement Models:
Replacement of Equipment fails Suddenly - Replacement policy when the value of money
changes with time - Group Replacement Models.
CHAPTER 5
Game Theory: Saddle point determination - Dominance property - Two Person zero sum
Game - Graphical Method with 3 x 3 Game - Simulation: Monte Carlo simulation
Application in Inventory Management and Queueing.
CONTENTS
CHAPTER I
1.0 Introduction
1.1 Formulation of lpp
1.2 Standard form of linear programming problem
1.3 Graphical method
1.3.1 Limitations of graphical method
1.4 Simplex Method
1.4.1 Solving linear programming problem using simplex method
1.4.2 Exercise problems
1.5 Duality in linear programming problem
1.6 Dual simplex method
1.7 Integer programming
1.8 Dynamic programming problem
CHAPTER IITRANSPORATION PROBLEMS
2.0 Definition
2.1 Mathematical formulation of transportation problem
2.1.1 Balanced transportation problems
2.1.2 Un balanced transportation problems
2.1.3 Methods for Finding initial basic feasible solution
2.2 North west corner method
2.3 Least Cost Method (or) Matrix Method
2.4 VOGELS approximation method (VAM)
2.4.1 Solving an un balanced transportation problem
2.5 Modified Distribution Method (MODI)
2.5.1 Procedure to find optimum solution (MODI Method)
2.6 Assignment Problem
2.7 Travelling Salesmen Problem
2.8 Transhipment Problem
2.8.1 Characteristics of transhipment problem
CHAPTER III NETWORK ANALYSIS
3.1 Definition of Network link and path
3.2 Minimum spanning tree
3.3 Shortest path problems (using Djikstra’s algorithm)
3.4 Maximum Flow problem
3.5 Sequencing Problem
3.6 Decision Theory
CHPTER IV QUEUEING THEORY
4.0 Queueing Theory concepts
4.1 Model I - (M/M/1) : (∞/FIFO)
4.2 Model II - (M/M/1) : (ࡺ/FIFO)
CHPTER V GAME THEORY
5.0 Game Theory Concepts
5.1 Odd’s Method
5.2 Matrix Oddment Method
5.3 Simulation
5.4 MONTE – CARLO Simulation
NOTES
MBAD 1923 QUANTITATIVE TECHNIQUES
CHAPTER I LINEAR PROGRAMING
STRUCTURE
1.0 Introduction
1.1 Formulation of LPP
1.2 Standard form of Linear Programming problem
1.3 Graphical method
1.3.1 Limitations of Graphical Method
1.4 Simplex Method
1.4.1 Solving Linear Programming problem using Simplex method
1.4.2 Exercise problems
1.5 Duality in linear programming problem
1.6 Dual Simplex method
1.7 Integer Programming
1.8 Dynamic Programming problem
1.0 INTRODUCTION
Definition
The linear programming problem(LPP) is to determine the values of the
decision variables such that all the constraints are satisfied and gives the
maximum or minimum value of the objective function.
1.1 FORMULATION OF LPP
Problem 1.
An egg contains 6 units of Vitamin A per gram and 7 units of Vitamin B per
gram and costs 12 paise per gram. Milk contains 8 , 12 units of Vitamin A and
Vitamin B respectively and costs 20 paise per gram. The daily minimum
1 SRMIST DDE Self Instructional Material
requirements of Vitamin A and Vitamin B are 100 units and 120 units
NOTES respectively. Find the optimal product mix. Formulate a LP model for the
above problem.
Solution. Let x and y be the no of units of egg and milk respectively. The
details of the product mix of the problem are given in the following table.
The LPP model of the problem is given below
Minimize z=12x+20y subject to
6x+8y≥100
7x+12y≥120
x≥0, y≥0.
Problem 2.
An electronic company manufactures two models of toys each by separate
production batch. The daily capacity of production of first model is 60 toys
and that of second model is 75 toy dolls. Each unit of first model uses 10
pieces of certain electronic component whereas each unit of second model
uses 8 pieces. Profit per unit of first and second models are Rs. 30 and Rs. 20
respectively. What should be the daily production of each model to get
maximum profit from the two models? Give a mathematical formulation to
the LPP.
Solution.
Let x and y be the number of units of production of two models of toys.
Constraint on capacity of first model is x≤60.
2 SRMIST DDE Self Instructional Material
Constraint on capacity of second model is y≤75
NOTES
Constraint on availability of electronic components for the two models is
10x+8y≤800.
Total profit on two models is z = 30x + 20y.
The non-negative constraints are x≥0, y≥0.
The mathematical formulation of the LPP is
Maximize z = 30x + 20y subject to
10x+8y≤800;
x≤60;
y≤75;
x≥0, y≥0.
Exercise problems.
1. A person requires 15, 13,14 units of chemicals A, B, C respectively for
his mango grove. These chemicals are available in liquid product in
jars and dry products in bags. The liquid product contains 6, 2, 2 units
of A, B, C respectively per jar. The dry product contains 2, 3, 4 units of
A, B, C respectively per bag. I the liquid product costs Rs. 8 per jar
and the dry product costs Rs. 3 per bag, how many of each should he
purchase in order to minimize the cost the requirements? Express this
problem as a LPP.
Answer. Minimize z =8x+3y subject to 6x+2y≥15, 2x+3y≥13, 2x+4y≥14, x≥0,
y≥0.
2. A company produces two types of pens A and B. Pen A is superior
quality and pen B is a lower quality. Profit on Pens A and B is Rs. 5
and Rs. 3 respectively. Raw materials required for each pen A is twice
as that of pen B. The supply of raw materials is sufficient only for 1000
pens of type B per day. Pen A requires a special clip and only 400 such
clips are available per day. For pen B only 700 clips are available per
day. Find the product mix so that the company can make maximum
profit. Prepare a linear programming model for this problem.
3 SRMIST DDE Self Instructional Material
Answer. Maximize z = 5x + 3y subject to 2x+y≤1000, x≤400,y≤700, x≥0, y≥0.
NOTES
3. An air force is experimenting with three types of bombs P, Q and R in
which three kinds of explosives A, B and C will be used. Taking the
various factors into consideration, it has been decided to use
maximum 600 kgs of explosive A, at least 480 Kgs of explosive B and
exactly 540 Kgs of explosive C. Bomb P requires 3,2,2 Kgs of A, B, C
respectively. Bomb Q 1, 4, 3 Kgs of A, B, C respectively. Bomb R
requires 6, 2, 3 Kgs of A, B, C respectively. Bomb P can exert a 2 tons
of explosion; Bomb Q can exert a 3 tons of explosion; Bomb R can
exert a 4 tons of explosion. What is the production schedule the air
force can make to have biggest bang. Write the mathematical model
for above problem.
Answer.Maximize z = 2x1 + 3x2+4x3subject to 3x1 + x2+ 6x3 ≤ 600, 2x1 + 4x2+
2x3≥480, 2x1 + 3x2+ 3x3 = 540, x1 ,x2 ,x3≥ 0.
GENERAL FORM OF LINEAR PROGRAMMING PROBLEM
Maximize/minimize z = c1x1+ c2x2+ . . . +cnxn ……………………..(1)
Subject to the following constraints
a11x1+ a12x2+ . . . +a1nxn(≤ = ≥) b1
a21x1+ a22x2+ . . . +a1nxn ( ≤ = ≥ ) b2 ......................... (2)
……………
am1x1+ am2x2+ . . . +amnxn ( ≤ = ≥ ) bm
and xi≥ 0 , i = 1, 2, ….. n. ………………...(3)
Here the x1, x2, …xn are called decision variables which are to be determined.
The function z = c1x1+ c2x2+ . . . +cnxnis called the objective function of an
LPP. c1, c2, …cnare called the cost coefficients. The conditions given by the
equation (2) are called constraints of the LPP.
The coefficients aij≥’ s are called the structural(technological) coefficients
and the coefficients bi’ s called stipulations(requirements).
The constraints xi≥0 ,i = 1, 2, ….. n are called the non-negativity constraints.
4 SRMIST DDE Self Instructional Material
NOTES
1.2 STANDARD FORM OF LINEAR PROGRAMMING PROBLEM
Definition – Slack variable
Consider a constraint of the form a11x1+ a12x2+ . . . +a1nxn ≤ b1. It is of ≤ type in
equation. This constraint can be expressed in an equation form by
introducing a new variable S1 ≥ 0 as follows.
a11x1+a12x2+ . . . +a1nxn +S1 = b1 where S1 ≥ 0 .
The new variable S1 is called slack variable.
Definition – Surplus variable
Consider a constraint of the form a11x1+ a12x2+ . . . +a1nxn ≥ b1. It is of ≥ type in
equation. This constraint can be expressed in an equation form by
introducing a new variable S1 ≥ 0 as follows.
a11x1+a12x2+ . . . +a1nxn - S1 = b1 where S1 ≥ 0 .
The new variable S1 is called surplus variable.
Hence using above definitions any LPP can be expressed in the following
form.
Maximize z = c1x1+c2x2+ . . . +cnxn
Subject to the following constraints
a11x1+a12x2+ . . . +a1nxn ± S1 = b1
a21x1+a22x2+ . . . +a1nxn ± S2= b2
……………
am1x1+am2x2+ . . . +amnxn ± Sm = bm
and xi≥ 0 , i = 1, 2, ….. n.
This is known as the standard form of LPP.
xi ,i = 1, 2, ….. n is said to be a feasible solution to an LPP if satisfies the
constraints and non negativity constraints.
5 SRMIST DDE Self Instructional Material
NOTES Definition – feasible solution
A set of values which satisfies the constraints and non-negativity restrictions
of the LPP is called the feasible solution of LPP.
Definition - optimal solution
A feasible solution which also optimize the objective function is called
optimal solution.
Definition - optimal basic feasible solution
A basic feasible solution solution which optimize the objective function is
called optimum basic feasible solution.
Definition – Degenerate solution
If there exists at least one basic variable whose value is zero then the basic
solution is said to be degenerate. Otherwise it is known as non-degenerate
basic solution.
1.3 GRAPHICAL METHOD
Procedure
Step 1. To use graphical method the number of variables in an LPP should be
less than are equal to 2.
Step 2. Draw all the constraints in the graph(by considering the inequalities as
equations).
Step 3. Find the common region satisfying all the constraints.
Step 4. Find the extreme points (corner points) of the common region.
Step 5. Substitute all the extreme points in the objective function. Choose the
point which gives maximum/minimum value to the objective function. This
point will be the solution for the LPP.
6 SRMIST DDE Self Instructional Material
1.3.1 Limitations of Graphical method NOTES
It is suitable for the LPP which is having only two variables.
More number of constraints in LPP leads to complication.
Problem 1. Solve the following linear programming problem graphically.
Maximize Z = 4x + y
Subject to the constraints:
x+y ≤ 50;
3x + y ≤ 90;
x ≥ 0, y ≥ 0.
Solution:
The non-negativity constraints x ≥ 0, y ≥ 0 implies that the solution region lies
on the first quadrant.
Consider the inequationx+y ≤ 50 as equation x+y = 50.
To draw this line we need two points on the line. Put x = 0 then we get y = 50
and put y = 0 then we get x = 50.
7 SRMIST DDE Self Instructional Material
So (0, 50) and (50, 0) are the two points lies on the line. Plot this point and
NOTES draw the line x+y = 50.
Consider the inequality 3x + y ≤ 90 as equation 3x + y = 90.
Identify two points on the line. Put x = 0 then we have y = 90. Put y = 0, we
get x = 30.
Plot the points and connect to draw the line 3x + y = 90 (observed from the
second line).
The shaded region in the above figure is the required feasible region
determined by the constraints x+y ≤ 50, 3x + y ≤ 90.
We observe the feasible region OABC bounded. So, we now use the corner
points to determine the maximum value of Z.
The coordinates of the corner points O, A, B and C are (0,0), (30, 0), (20, 30)
and (0, 50) respectively. Now we evaluate the Z at each corner point
Hence the maximum value of Z is 120 at the point (30, 0).
The solution for the LPP is x = 30 ; y = 0 ; and z = 120 .
Problem 2. Solve the following Linear programming problem graphically.
Minimize Z = 200 x + 500 y
Subject to the constraints:
x + 2y ≥ 10,
3x + 4y ≤ 24,
x ≥ 0, y ≥ 0.
8 SRMIST DDE Self Instructional Material
Solution: The shaded region ABC in the following figure is the required
NOTES
region by the system of constraints.
The non-negativity constraints x ≥ 0, y ≥ 0 implies that the solution region lies
on the first quadrant.
Consider the in equation x + 2y ≥ 10, as equation x + 2y = 10,
To draw this line we need two points on the line. Put x = 0 then we get y = 5
and put y = 0 then we get x = 10.
So (0, 5) and (10, 0) are the two points lies on the line. Plot this point and
draw the line x + 2y = 10,.
Consider the inequality 3x + 4y ≤ 24 as equation 3x + 4y = 24.
Identify two points on the line. Put x = 0 then we have y = 6. Put y = 0, we get
x = 8.
Plot the points and connect to draw the line 3x + 4y = 24 (observed from the
second line).
The shaded region in the above figure is the required feasible region
determined by the constraints x + 2y ≥ 10, 3x + 4y ≤ 24.
9 SRMIST DDE Self Instructional Material
We observe the feasible region ABC bounded. So, we now use the corner
NOTES points to determine the maximum value of Z.
The coordinates of the corner points A, B and C are (0, 5), (4, 3) and (0, 6)
respectively. Now we evaluate at these points.
Hence, minimum value of Z is attained at the point (4, 3).
The solution for the LPP is x = 4; y = 3; and z = 2300 .
Problem 3. Determine graphically the minimum value of the objective
function Z = -50 x + 20 y
Subject to the constraints:
2x – y ≥ -5,
3x + y ≥ 3,
2x – 3y ≤ 12,
x ≥ 0, y ≥ 0.
Solution.
The following diagram shows the feasible region obtained from the
constraints.
10 SRMIST DDE Self Instructional Material
NOTES
The feasible region shown in the figure is unbounded. Now evaluate Z at the
corner points.
From the above table we find that -300 is the smallest value of Z at the corner
point (6, 0).
Can we say that the minimum value of Z is -300 ?. Note that if the region
would have been bounded, this smallest value of Z is the minimum value of Z.
But here we see that the feasible region is unbounded. Therefore, -300 may or
may not be the minimum value of Z. To decide this issue, we graph the
inequality-50x + 20y < -300 and check whether the resulting open half plane
11 SRMIST DDE Self Instructional Material
has points in common with feasible region or not. If it has common points,
NOTES then -300 will be the minimum value of Z.
In the above example, can you say whether Z = -50 x + 20 y has the maximum
value 100 at (0, 5) ? For this, check whether the graph of -50 x + 20 y > 100 has
points in common with the feasible region.
1.4 SIMPLEX METHOD
1.4.1 Solving Linear programming problem using simplex method
Procedure:
Step 1. Express the given LPP in to the standard form by introducing slack or
surplus variables.
Step 2. Find the initial basic feasible solution.
Step 3. Test for the optimality
Case 1. If Zj– Cj≥ 0 for all j. In this case the current basic feasible solution is the
required optimal solution.
Case 2. If Zk– Ck< 0 for some k and all the corresponding column values are
negative(aik< 0) for all i = 1, 2, … ,m. In this case the solution is unbounded
solution.
Case 3. If Zj– Cj< 0 and there exists at least one i such that aij≥ 0. We choose j
such that Zj– Cjis minimum (most negative).
If there is tie between Zj– Cj‘s, it is broken by the arbitrary choice. This
determines a variable that is currently non-basic and becomes a basic variable
in the next iteration.
For this reason, xj is called the entering variable and the correspondingjth
column to be entered is called the key column.
Step 4. Calculate the ratio (XBcolumndivided by the entering variable
column). Find the minimum ratio that corresponding row will be the leaving
variable row.
12 SRMIST DDE Self Instructional Material
Step 5. The intersection point of the leaving variable row and the entering
NOTES
variable column is called the pivotal element or key element.
Step 6. Do the first iteration replace the leaving variable by the entering
variable.
Divide all the elements of the row by the key element and the remaining rows
are obtaining by the elementary row operations, so as to make the all the
entries in the column corresponding to the entering variable zero.
Using the following formula we get the new values of the table.
Thus we get the new iteration. Do the optimality checking again.
Repeat the process until we reach the optimality.
Problem 1.
Solve the following LPP by simplex method.
Maximize Z = 25x1+20x2subject to the following constraints
16x1 + 12x2 ≤ 100,
8x1 + 16x2 ≤ 80,
x1≥0, x2≥0.
Solution:
We rewrite the LPP in standard form by introducing the slack variables s1, s2
to the left hand side of the inequalities.
16x1 + 12x2 +s1= 100,
8x1 + 16x2 +s2 = 80,
x1, x2 , s1 and s2≥0.
s1and s2 are the initial basic variable.
Prepare the initial simplex table and test for optimality. If the table is not
optimum then improve the table to reach optimum solution.
13 SRMIST DDE Self Instructional Material
NOTES
In theinitial simplex table two values of Zj– Cjare negative. Hence the initial
basic feasible solution is not optimal. So we go to next iteration.
Entering variable:We choose the most negative of these two net evaluations
Z1 – C1 and Z2 – C2minimum of (-25, -20) = -25 that corresponds to column x1.
Since all the components of x1 are positive and x1 is the entering variable. The
new column x1 is called the key column.
Leaving variable:To determine the leaving variable we find the ratio by
using the formula ‘XB columndivided by the entering variable column’.
The minimum value of the ratio is 6.25. Hence, the corresponding variable s1
is the leaving variable.
Key element: The intersection of entering variable column and leaving
variable row is 16. Thus 16 is the key element.
First iteration: We replace the basic variable s1 by x1. Divide all the elements
in the key row by 16. The other rows of the new table filled by using the
following formula,
For example, to find the new value of 16 in second row second column of the
cost matrix in the first iteration table, we have
14 SRMIST DDE Self Instructional Material
New value = 16 – ((8/16) x 12 ) = 16 – 6 = 10.
NOTES
Thus we get the following table.
In the first iteration simplex table one of Zj– Cjis negative. Hence the initial
basic feasible solution is not optimal. So we go to next iteration. By repeating
the above process we reach the second iteration table.
All the Zj– Cjare positive in the above table. So, we reach the optimality.
The optimal solution is
x1= 4 ; x2 = 3 and
The value of the objective function
15 SRMIST DDE Self Instructional Material
Z = 25x1+20x2 = 25 x 4= 25 x 4+ 20 x 3 = 160.
NOTES
Problem 2. Maximum Z = 30x1+20x2, where x1≥0, x2≥0 subject to the
following constraints
10x1 + 8x2≤ 800;
x1≤ 60;
x2≤ 75;
x1≥0, x2≥0.
Solution: Add the slack variables s1, s2 and s3 to the left hand side of the
inequalities to obtain the standard form of the LPP.
10x1 + 8x2 +s1 = 800;
x1 +s2 = 60;
x2+ s3 = 75;
One of Zj– Cjis negative so the current table is not optimal. x1is the entering
variable and s2 is the leaving variable; 1 is the pivotal element. Here the ratio
for s3 does not exist and hence it is left out in ratio column.
16 SRMIST DDE Self Instructional Material
NOTES
One of Zj– Cjis negative so the we improve the current table to reach
optimality. x2 is the entering variable and s1 is th leaving variable. 8 is the key
element.
All the Zj– Cjare positive so the optimality attained. The optimal solution is x1
= 60; x2 = 25 and the maximum value of z is 2300.
17 SRMIST DDE Self Instructional Material
1.4.2 Exercise problems.
NOTES
Solve the following Linear programming models using graphical method and
simplex method.
1. Maximize Z = 2x1+x2, subject to the following constraints
x1+ 2x2 ≤ 10
x1+ x2≤ 6
x1- x2 ≤ 2
x1- 2x2 ≤ 1,
x1≥0, x2≥0.
Solution: x1 = 2, x2 = 1 and max Z = 10.
2. Maximize Z = 20x1+40x2, subject to the following constraints
5x1 + 10x2 ≤ 50
10x1 + 5x2≤ 50
x1≥0, x2≥0.
Solution: x1 = 10/3, x2 = 10/3 and max Z = 200; x1 = 0, x2 = 5 and max Z = 200
(Two optimal solution for this problem)
3. Maximize Z = 3x1+2x2, subject to the following constraints
x1- x2 ≤ 1
3x1 - 2x2≤ 6
x1≥0, x2≥0.
Solution: Unbounded solution.
4. Minimize Z = 4x1+3x2, subject to the following constraints
2x1 - 3x2 ≥ 6
6x1 + 5x2 ≥ 30
x1≥0, x2≥0.
Solution: x1 = 0, x2 = 6 and min Z = 18.
18 SRMIST DDE Self Instructional Material
1.5 DUALITY IN LINEAR PROGRAMMING PROBLEM NOTES
The canonical form of a Linear Programming Problem is
Maximize z = c1x1+ c2x2+ . . . +cnxn
Subject to the following constraints
a11x1+ a12x2+ . . . +a1nxn ≤ b1
a21x1+ a22x2+ . . . +a1nxn ≤ b2
…………………..
am1x1+ am2x2+ . . . +amnxn ≤ bm
and xi ≥ 0 , i = 1, 2, ….. n. This is called a primal problem and the variables are
called primal variables and the constraints are called primal constraints.
The dual of the LPP is defined by
General form of a Linear Programming Problem
Minimize z = b1w1+ b2w2+ . . . +bnwn
Subject to the following constraints
a11w1+ a12w2+ . . . +a1nwn≥ c1
a12w1+ a22 w2+ . . . +a1nwn≥ c2
……………
a1nw1+ a2nw2+ . . . +amnwn≥cn
andwi ≥ 0 , i = 1, 2, ….. n.
The variables wis are called dual variables and the constraints are called the
dual constraints.
Problem 1. Form the dual of the following primal LPP.
Maximize Z = 4x1+10x2+25x3, subject to the following constraints
2x1+4x2+8x3≤ 25
4x1+9x2+8x3≤ 30
x1, x2 , x3≥0.
19 SRMIST DDE Self Instructional Material
Solution: Minimize W = 25w1+30w2 subject to
NOTES
2w1+4w2≥ 4,
4w1+9w2≥ 10,
8w1+8w2≥ 25, w1, w2 ≥ 0
Problem 2. Form the dual of the following primal LPP.
Maximize Z = x1+2x2+3x3 , subject to the following constraints
4x1+5x2+4x3 ≤ 9
6x1- x2+5x3 =10
x1, x2 , x3≥0.
Solution: Minimize W = 9w1+10w2 subject to
4w1+6w2≥ 1,
5w1-w2≥ 2,
4w1+5w2≥ 3,
w1 ≥ 0 and w2 is unrestricted.
20 SRMIST DDE Self Instructional Material
1.6 DUAL SIMPLEX METHOD NOTES
Step 1. Convert the given problem to maximization type if it is in
minimization type. Convert the ≥ type constraints in to ≤ type constraints by
multiplying both side of such constraints by -1.
Step 2. Introduce the slack variables and obtain an initial basic feasible
solution by regular simplex method.
Step 3.(i) If all Zj-Cj≥ 0 and XB ≥ 0 then the current solution is optimal
solution.
(ii) If all Zj-Cj≥0 and at least one basic variable XB≤ 0 then go to step 4.
(iii) If any one of Zj-Cj≤ 0 then this method cannot be applied to the
given problem.
Step 4.(i) Select the most negative XB. Then the corresponding variable is the
leaving variable.
(ii) If all aij≥ 0 the there is no feasible solution for the problem.
Zj−Cj
(iii) If any aij≤ 0 the compute the ratios , 𝑎𝑖𝑗 < 0. choose the
aij
maximum ratio then the corresponding variable is the entering variable.
Step 5. Using the pivotal element do the next iteration and do the step 3 until
we reach the feasible solution.
Problem 1.Solve the following LPP by dual simplex method
Minimize z = 60 x1+ 80 x2 subject to
20 x1+ 30 x2≥ 900 , 40 x1+ 30 x2≥ 1200,x1, x2 ≥ 0.
Solution. The LPP is of minimizing type. Hence, it can be converted into
maximization type as
Maximize z* = - 60 x1- 80 x2 subject to
-20 x1- 30x2+s1 ≤ -900 , -40x1- 30x2+s2 ≤ -1200
x1, x2 , s1, s2 ≥ 0.
21 SRMIST DDE Self Instructional Material
NOTES
Since all Zj*- Cj≥ 0, optimal solution is obtained but XB ≤ 0 so the solution is
not feasible. To obtain the optimum feasible solution we go for a new basis.
The most negative XB is -1200 which corresponds to the variable s2 .
Zj−Cj
To obtain the entering variable calculate the ratio . The maximum of the
aij
ratio is -3/2 . The corresponding variable x1 is the entering variable.
22 SRMIST DDE Self Instructional Material
In the above table the most negative XB is -300 and the maximum ratio is
NOTES
-35/15. So x2 enters and s1 leaves the basis. The new improved table follows.
From the above table, we get
Max z* = -2500. We know that Min z = - Max z* = - (-2500 ) = 2500.
The optimum feasible solution is x1 = 15 ; x2 = 20 and min z = 2500.
Exercise problems.
Problem1. Using the dual simplex method solve the LPP.
Maximize z = - x - y subject to
2 x+ y ≥ 2 , - x- y ≥ 1, x, y ≥ 0.
Solution: No feasible solution.
Problem 2. Using the dual simplex method solve the LPP.
Maximize z = - 3x - y subject to
x+ y ≥ 1 , 2 x+3y ≥ 2, x, y ≥ 0.
Solution: x = 0 ; y = 1 and max z = -1 .
23 SRMIST DDE Self Instructional Material
1.7 INTEGER PROGRAMMING
NOTES
The linear programming problem with the additional requirements that the
variables can take on only, integer values may have the following
mathematical form.
Maximize or minimize z = c1x1+c2x2+ . . . +cnxn
Subject to the following constraints
a11x1+a12x2+ . . . +a1nxn = b1
a21x1+a22x2+ . . . +a1nxn = b2 ≤
…………………..
am1x1+am2x2+ . . . +amnxn= bm,
xi≥ 0 ,i = 1, 2, ….. n and xi are integer valued for i = 1,2,….p≤ n.
Note.
1. An integer programming problem in which all variables are required to
be integers is called a pure or all-integer programming problem. For
example, the following LPP is a pure integer programming problem.
Maximize z = 9x1+10x2 subject to the following constraints
4x1+3x2 ≥ 4, x1 ≤ 4, x2 ≤ 8 where x1 ≥ 0, x2 ≥ 0 and are integers.
2. An integer programming problem in which only some of the variables
are required to be integers is known as a mixed integer programming
problem. For example, the following LPP is a mixed integer
programming problem.
Maximize z = 4x1-8x2 subject to the following constraints
-x1+3x2 ≤ 7, x1 + x2 ≤ 4, x1 - 4x2 ≤ 9 where x1 ≥ 0, x2 ≥ 0 and x2 is an integer.
3. An integer programming problem in which all the variables must have
integer values obly zero or unity is called the zero-one integer
programming problem.
24 SRMIST DDE Self Instructional Material
Gomory’s fractional cut method or cutting plane method: NOTES
An iterative procedure for the solution of an all integer linear
programming problem by Gomory may be summarized in the
following steps
Express the given LPP into the standard form.
Determine an optimum solution by using simplex method ignoring
the integer value restriction.
Test the integrality of the optimum solution.
If the optimum solution admits all-integer values, an optimum basic
feasible solution is attained.
If the optimum solution does not include all integer values, then move
on to next step.
Choose the largest fractional value of the basic variables. Let it be fk.
Express each of the negative fractions, if any, in the kth row of the
optimum simplex table as the su of a negative integer and a non-
negative fraction
Generate the Gomorionconstraint(fractional cut) in the form G1 = - fk +
fk1 x1 + fk2 x2 + . . . +fknxn where 0<fkj< 1 and 0 ≤ fk< 1.
Add the Gomorian constraint at the bottom of the optimum simplex
table.
Use dual simplex method to find an improved optimum solution.
Repeat the procedure until an optimum basic feasible all integer
solution is obtained.
Problem 1. Solve the following integer programming problem.
Maximize z = 2x+3y subject to the following constraints
2x+2y ≤ 7,
x≤ 2,
y ≤ 2 where x, y≥ 0 and are integers.
25 SRMIST DDE Self Instructional Material
Solution.
NOTES
Introduce the slack variable s1, s2 ≥ 0 and use the simplex method to find the
optimum basic feasible solution for the given IPP. The last iteration of the
simplex table is given below
Using the simplex method we got the above optimum table. From the table
we can observe the optimum solution is
x = 3/2 and y = 2
The solution of x is not an integer to get the integer solution for the given
integer programming problem, we should find the fractional cut and use
Gomorian algorithm to reach integer solution for the variables.
The solutions x and y can be expressed as
x = 1+ ½ the fractional part of x represented by fx and fx= 1/2 .
y = 2+ 0 the fractional part of y represented by fyandfy= 0.
Take the maximum of (fx ,fy ) = maximum of (1/2 , 0 ) = ½.
26 SRMIST DDE Self Instructional Material
This maximum value 1/2 is corresponding to the variable x. So we should
NOTES
use the row corresponding to x in the above simplex table to frame the new
equation.
1 x + 0 y + ½ s1+ 0 s2 – 1 s3 = 3/2
Now rewrite all the coefficients in the above equation into integer+fraction
format. Then we get
(1+0) x + (0+0) y +(0 + ½ ) s1 + (0 + 0) s2 – ( 1+ 0) s3 = (1 + ½ )
Considering only the fractional part of the coefficients in the above equation,
we get
(1+0) x + (0+0) y +(0 + ½ ) s1 + (0 + 0) s2 – ( 1+ 0) s3 = (1 + ½ )
this implies 0 x + 0 y + ½ s1 + 0 s2 – 0 s3 = ½
this implies ½ s1 = ½
The fractional cut must be a greater than or equal to constraint so rewrite
above equality as
½ s1 ≥ ½
this implies - ½ s1 ≤ - ½
Now introduce the Gomorian slack variable G ≥ 0 to make this inequality as
equality
- ½ s1 + G = - ½
Add the above constraint in the simplex table and check the optimality again.
27 SRMIST DDE Self Instructional Material
NOTES
In the above table we estimated the Zj-Cj value all are positive but the one of
the solution value is negative( G1 = -½ ).
So we should use the dual simplex method to find the positive solution.
The solution of the G1 is negative so the pivotal row is the row corresponding
to G1.
Find the ratio and choose the maximum of this ratio. Then the pivotal
element is -½ .
The next iteration follows
28 SRMIST DDE Self Instructional Material
All Zj-Cjare positive and all the corresponding solutions also positive and
NOTES
integer we reach the optimum integer solution.
The solution for the given integer programming problem is
x = 1 ; y = 2 and maximum of z = 8.
Exercise problem
1. Find the optimum integer solution to the following LPP
Maximize z = x+4y subject to the following constraints
2x+ 4y ≤ 7,
5x +3y ≤ 15 where x, y ≥ 0 and are integers.
Solution: x = 1; y = 1 and z = 5.
2. Solve the IPP using cutting plane method.
Maximize z = 3x1+ x2+3x3subject to the following constraints
-x1+ 2x2+ x3 ≤ 4 ,
4 x2- 3x3 ≤ 2,
x1- 3x2+2x3 ≤ 3 where x1, x2, x3 ≥ 0 and all are integers.
Solution: x1 = 5; x2 = 2; x3 = 2 and z = 23.
3. Solve the IPP using cutting plane method
Maximize z = x1 - 2x2 subject to the following constraints
4x1+ 2x2≤ 15 where x1, x2 ≥ 0 and all are integers.
Solution: x1 = 3; x2 = 0 and z = 3.
29 SRMIST DDE Self Instructional Material
1.8 DYNAMIC PROGRAMMING PROBLEM
NOTES
Definition.
Dynamic programming is a mathematical technique dealing with the
optimization of multistage decision problem
Dynamic programming differs from LPP in two ways
In DPP a standard mathematical formulation does not exists. Particular
equation must be developed to fit for each individual situation.
LP gives one time solution where as DP considers decision making over time
and solves each sub-problems optimally.
DPP is a technique for solving problems in which decision have to be made
each successive stages. It provides a systematic procedure for determining the
combination of decision that maximizes overall effectiveness.
In DPP, the problem is divided into number of sub problems and the
solutions of sub problems are then combined to get overall solution.
Definition.
Stage. The DPP is broken into sub problems and each sub problem is called
stage.
Stage Decision. At each stage there are number of alternatives and the
selection of one of the most suitable and feasible alternative is called stage
decision.
Definition.
State variables. The variable whose values specify the condition of the
decision process and summarize the current status of the system are called
state variables.
Definition.
Decision variables. The unknown in the given problem that need to be
determined are called decision variables.
Definition.
Return function. At each stage, a decision is made which can affect the state
of the system at the next stage. It can be described in an algebraic equation
form. The equation which represents the contribution of each stage is called a
return function.
30 SRMIST DDE Self Instructional Material
NOTES
31 SRMIST DDE Self Instructional Material
Definition.
NOTES Recursive relationship. It represents mathematical expression connecting the
optimal solution of the earlier stage sand the return function from the current
stage.
Dynamic Programming Algorithm
Identify the decision variables and specify objective function to be
optimized under certain limitations if any.
Divide the given problem into a number if smaller stages.
Identify the state variables at each stage and write down the
transformation function as a function of state variable and decision
variable at the next stage.
Write down the general recursive relationship for computing the
optimal policy.
Decide whether forward or backward method is to follow to solve the
problem
Construct appropriate stages to show required value of the return
function at each stage.
Determine the overall optimal policy or decisions and its value at each
stage.
Forward and Backward computational procedure
When the DPP is solved using the recursive equation starting from the
1st stage to last stage the computation involved is called the forward
computational procedure.
Forward and Backward computational procedure
When the DPP is solved using the recursive equation starting from the
last stage to first stage the computation involved is called the
backward procedure.
32 SRMIST DDE Self Instructional Material
CHAPTER II NOTES
TRANSPORTATION ASSIGNMENT AND TRANS-SHIPMENT
Structure
2.0 Definition
2.1 Mathematical formulation of transportation problem
2.1.1 Balanced transportation problems
2.1.2 Un balanced transportation problems
2.1.3 Methods for Finding initial basic feasible solution
2.2 North west corner method
2.3 Least Cost Method (or) Matrix Method
2.4 VOGELS approximation method (VAM)
2.4.1 Solving an un balanced transportation problem
2.5 Modified Distribution Method (MODI)
2.5.1 Procedure to find optimum solution (MODI Method)
2.6 Assignment Problem
2.7 Travelling Salesmen Problem
2.8 Transhipment Problem
2.8.1 Characteristics of transhipment problem
33 SRMIST DDE Self Instructional Material
2.0 DEFINITION
NOTES
A transportation problem determines a minimum cost plan to transport a
homogenous commodity from various sources destination.
2.1 Mathematical formulation of transportation problem
𝑚 𝑛
Minimize 𝑧 = 𝑗=1 𝐶𝑖𝑗 𝑋𝑖𝑗 subject to
𝑖=1
𝑋𝑖𝑗 ≤ 𝑎𝑖 i=1,2,3…..n (supply constraints)
𝑋𝑖𝑗 ≥ 𝑏𝑗 j=1,2,3…..n (demand constraints)
Xij ≥ 0; i=1,2,3…..n and j=1,2,3…..m (non-negativity constraints)
2.1.1 Balanced transportation problems
If the total demand is equal to total supply then the transportation problem is
called balanced transportation problem
𝑚 𝑛
(ie) 𝑖=1 𝑎𝑖 = 𝑗 =1 𝑏𝑗.
2.1.2 Un balanced transportation problems
If the total demand is not equal to total supply then the transportation
problem is called balanced transportation problem
𝑚 𝑛
(ie) 𝑖=1 𝑎𝑖 ≠ 𝑗 =1 𝑏𝑗.
2.1.3 Methods for Finding initial basic feasible solution
Northwest corner method
Least cost method (or) Matrix minima method
Vogel approximation method.
34 SRMIST DDE Self Instructional Material
2.2 NORTH WEST CORNER METHOD NOTES
Problem 1. Using North West corner rule find the initial basic feasible
solution to the following transportation problem.
Solution
𝑎 𝑖 = 900 + 1000 + 1200 = 3100&
𝑏𝑗 = 1200 + 1000 + 900 = 3100
So , the given TP is a balanced problem.
Step 1:
Identify the northwest corner cell and compare the corresponding supply and
demand.
In our following table the northwest corner cell is 8.
Allocate the minimum of {supply , demand} to the cell then eliminate the
corresponding row or column.
In our problem , min {900,1200}=900 so, eliminate the first row.
Also subtract 900 from 1200 and change the demand for the corresponding
column as 1200 - 900 = 300.
35 SRMIST DDE Self Instructional Material
NOTES
Step 2:
Repeat the above process,we get min{1000,300}=300 so allocate 300 to the
north west celland subtract 300 from supply 1000.
Eliminate corresponding column.
Step 3
Minimum of {700,1000}=700 so allocate 700 to the north west cell.
Eliminate corresponding row also subtract the 700 from the demand 1000.
Then the next iteration will be
36 SRMIST DDE Self Instructional Material
NOTES
Step 4.
Min{300,1200}=300 so allocate 300 and eliminate the column.
Subtract the 300 from the supply 1200
37 SRMIST DDE Self Instructional Material
NOTES
The basic feasible solution is
x11 = 900 ;
x21 =300 ;
x22 = 700 ;
x32 = 300 ;
x33 =900.
The total transportation cost is
900×8+12×300+13×700+10×300+11×900 = 32800
2.3 LEAST COST METHOD (OR) MATRIX METHOD
Problem1.
Use the least cost method to find the initial basic feasible solution to the
following transportation problem.
38 SRMIST DDE Self Instructional Material
NOTES
Solution.
The sum of all supply
𝑎 𝑖 = 900 + 1000 + 1200 = 3100
The sum of all demad
𝑏𝑗 = 1200 + 1000 + 900 = 3100
𝑚 𝑛
𝑖=1 𝑎𝑖 = 𝑗 =1 𝑏𝑗. So , the given TP is a balanced problem.
Select the smallest element in the cost matrix. In our problem the smallest in
the matrix is 8.
Compare the corresponding demand and supply . Allocate the , minimum
value to the cell.
Min{ 900,1000 }=900, so allocate 900 to the cell having minimum cost 8 and
eliminate the corresponding row.
The next iteration as follows
39 SRMIST DDE Self Instructional Material
NOTES
Now the minimum cost in the table is 10 .compare the corresponding supply
and demand 1200 min{ 1200,1000 }=1000.
So, allocate 1000 to the cell having the least cost 10 and eliminate the
corresponding row or column.
Now the minimum cost is 11.compare the corresponding demand 900 with
supply 200.
Min{ 900,200 }= 200 eliminate corresponding row.
40 SRMIST DDE Self Instructional Material
Now the minimum cost is 12 (if tie occurs we can select any one arbitrarily)
NOTES
Transportation cost = 900×8+300×12+700×12+1000×10+200×11= 31400 .
Basic feasible solution: x11=900 ; x21=300 ; x23=700 ; x32=1000 ; x33=200 .
41 SRMIST DDE Self Instructional Material
2.4 VOGELS APPROXIMATION METHOD (VAM)
NOTES
Problem1. Use the VAM to find initial basic feasible solution to the TP.
Solution
Solution.
The sum of all supply
𝑎 𝑖 = 900 + 1000 + 1200 = 3100
The sum of all demand
𝑏𝑗 = 1200 + 1000 + 900 = 3100
𝑚 𝑛
𝑖=1 𝑎𝑖 = 𝑗 =1 𝑏𝑗. So , the given TP is a balanced problem.
Step 1:
The difference between the least cost and next to least cost in each row known
as row penalty.
The difference between the least cost and next to least cost in each column
known as column penalty.
42 SRMIST DDE Self Instructional Material
NOTES
Select the largest value among penalties. Go to the corresponding row or
column.
In that row or column select minimum cost and made allocation.
In the above table largest penalty is 4 then select the first column in that first
column minimum cost is 8.
Find the minimum of {900, 1200} and allocate to 8.
Eliminate the row and subtract 900 from demand 1200.
43 SRMIST DDE Self Instructional Material
Step 2:
NOTES
The largest among penalties is 3 so select the second column .
In that second column the minimum cost is 10 .so compare the supply &
demand and allocate the minimum .
Allocate minimum {1200,1000} =1000 and subtract 1000 from the supply 1200.
Step 3
44 SRMIST DDE Self Instructional Material
The largest penalty is 2 so select the first column and identify the minimum
NOTES
cost 12.
Now find the minimum of { 300,1000 } =300 , allocate 300.
Delete the corresponding column and substract300 from the supply 1000.
Highest penalty is 12. So allocate 700 to the cell contains 12 and subtract the
700 from the supply 900.
45 SRMIST DDE Self Instructional Material
NOTES
Initial basisc feasible solution of the given transportation problem.
x11=900 ;
x21=300 ;
x23=700 ;
x32=1000 ;
x33=200.
46 SRMIST DDE Self Instructional Material
Transportation cost:
NOTES
900×8+300×12+700×12+1000×10+200×11= 31400.
Problem2.
Use Vogel approximate method and find the initial basic feasible
solution for the following transportation problem.
D1 D2 D3 D4 supply
S1 21 16 25 13 11
S2 17 18 14 23 13
S3 32 27 18 41 19
demand 6 10 12 15
Solution.
The sum of all supply 𝑎 𝑖 = 11 + 13 + 19 = 43.
The sum of all demand 𝑏𝑗 = 6 + 10 + 12 + 15 = 43.
𝑚 𝑛
𝑖=1 𝑎𝑖 = 𝑗 =1 𝑏𝑗. So , the given TP is a balanced problem.
The transportation problem is balanced so we can proceed further. The initial
basic feasible solution for the given problem is
47 SRMIST DDE Self Instructional Material
NOTES
The initial basic feasible solution is given by
x11=11 ;
x21=6 ;
x23=3 ;
x24=4 ;
x32=7 ;
x33=12
Transportation cost
11×13+6×17+3×18+4×23+7×27+12×18 = 796.
2.4.1 Solving an un balanced transportation problem
Problem1.
Convert the unbalanced transportation in to balanced transportation problem
and initial basic feasible solution by North west corner method .
D1 D2 D3 ai
S1 2 3 4 700
S2 4 2 3 900
S3 1 2 3 1000
Bj 400 500 800
48 SRMIST DDE Self Instructional Material
Solution.
NOTES
The sum of all supply 𝑎𝑖 = 700 + 900 + 1000 = 2600.
The sum of all demand 𝑏𝑗 = 400 + 500 + 800 = 1700 .
𝑚 𝑛
𝑖=1 𝑎𝑖 ≠ 𝑗 =1 𝑏𝑗. So , the given TP is a balanced problem .
The overall demand is not equal to supply so add a dummy column
with 0 cost and demand as 900, we have
The sum of all supply 𝑎𝑖 = 700 + 900 + 1000 = 2600.
The sum of all demand 𝑏𝑗 = 400 + 500 + 800 + 900 = 2600 .
𝑚 𝑛
𝑖=1 𝑎𝑖 = 𝑗 =1 𝑏𝑗. Now the TP becomes a balanced problem .
Using the northwest corner rule we reach the following table.
x11=400 ; x12=300 ; x22=200 ; x23=700 ; x33=100 ; x34=900
Transportation cost is
2×400+3×300+2×200+3×700+3×100+0×900= 4500.
49 SRMIST DDE Self Instructional Material
2.5 MODIFIED DISTRIBUTION METHOD (MODI)
NOTES
2.5.1 Procedure to find optimum solution (MODI Method)
Step 1: Find the Initial basic feasible solution by anyone of the above method
(preferably VAM)
Step2: Cells with allocations are called basic cells and cells without
allocations are non–basic cells. From basic cells frame a set of equations
Ui +Vj = Cij, whereCij’s are cost of basic cells
Step 3: Assume any one of Ui’s (or) Vj’s as zero and solve the equations to
find remaining Ui’s and Vj’s values
Step 4: Now estimate Ui + Vj- Cijfor all non – basic cells
Step 5: If all Ui + Vj– Cij’s are negative for all non – basic cells. Then the
current basic solution is the required optimum solution.
If any one of the Ui + Vj– Cij’s value is positive then we should improve the
current basic solution to reach optimality.
Step 6: To improve the solution draw a loop whose starting and ending point
should be a non-basic cell and corners of the loop should be a basic cell
Step 7: Let the starting point of the loop be ‘+θ’ and name the other corners
of the loop –θ, +θ alternatively .
Step 8: Find the minimum allocation value among the basic cells with ‘-‘sign
in the loop. That minimum allocation value is our θ.
Step 9: Add θ to the allocations if the cell is having ‘+’ sign and subtract θ if
the cell is having ‘-‘ ive sign. Now, we reach the improved basic solution.
Step 10: Now go to Step 2 & Step 3 to check the optimality of the improved
basic solution. Repeat the above process until you reach optimality.
50 SRMIST DDE Self Instructional Material
Problem 1: NOTES
Solve the Transportation problem
Solution:
Using VAM method we obtain the initial basic feasible solution as follows
The initial basic feasible solution is
x11 = 3 ;
x13 = 3 ;
x21 = 1 ;
x22 = 7.
51 SRMIST DDE Self Instructional Material
Transportation Cost is
NOTES
3 x 3 + 3 x 3 + 1 x 2 + 7 x 3 = 41
From basic cells we can frame the following equations
U1 +V1 = 3 ;
U1 +V3 = 3 ;
U2 +V1 = 2 ;
U2 +V2 = 3 ;
Put U1= 0, We get V1=3; V= 3 ;U2= -1 ; V2= 43
From non basic cells calculate
U1 +V2 – C12= 0 + 4 + 7 = -3
U2 +V3 – C23 = -1 + 3 – 9 = -7
All Ui + Vj– Cij’s are negative so current basic solutions is the optimum
solution
X11 = 3 ;
X13 =3 ;
X21 = 1 ;
X22 = 7
The optimum transport cost is
Z=3 x 3 + 3 x 3 + 1 x 2 + 7 x 3 = 41.
Problem 2:
Solve the transportation problem. Use Northwest corner rule to find the
Initial basic feasible Solution
52 SRMIST DDE Self Instructional Material
NOTES
Solution.
The sum of all supply
𝑎 𝑖 = 900 + 1000 + 1200 = 3100
The sum of all demad
𝑏𝑗 = 1200 + 1000 + 900 = 3100
𝑚 𝑛
𝑖=1 𝑎𝑖 = 𝑗 =1 𝑏𝑗. So , the given TP is a balanced problem.
Using the northwest corner rule we reach the following table.
53 SRMIST DDE Self Instructional Material
The basic Solution is
NOTES
x11 = 900 ;
x21 = 300 ;
x22 = 200 ;
x32 = 300 ;
x33 = 900
Transportation cost is 32800.
From basic cells, we get
U1 +V1 = 8
U2 +V1 = 12
U2 +V2 = 13
U3 +V3 = 10
U3 +V3 = 11
Put U1= 0, then we get V1=8; U2 = 4 ; V2= 9; U3 = 1; V3= 10
From the non basic cells the net evaluations are calculated as follows
U1 +V2 – C12 = 0 + 9 – 10 = -1
U1 +V3 – C13 = 0 +10 – 12 = -2
U2 +V3 – C23 = 4 +10 – 12 = 2
U3 +V1 – C31 = 1 + 8 – 14 = -5
Since U2 +V3 – C23 is positive. So we should improve the basic solution to
reach optimality.
The maximum positive Ui + Vj– Cij’s is occurred at the cell (2,3) form a loop
whose starting and ending point is the non basic cell (2,3).
54 SRMIST DDE Self Instructional Material
NOTES
θ= min { 900, 700} = 700. So add θ = 700 if the cell having +ive sign and
subtract if the cell having –ive sign.
900
8 10 12
300 700
12
13 12
14 1000 200
11
10
From basic cells, we observe
U1 +V1 = 8;
U2 +V1 = 12;
U2 +V3 = 12;
U3 +V3 = 1 ;
U3 +V2 = 10;
Put U1= 0, we get V1=8; U2 = 4 ; V3= 5; U3 = 3; V2= 7
From the non basic cells, the net evaluations are calculated as follows
55 SRMIST DDE Self Instructional Material
U1 +V2 – C12 = 0 + 7 – 10 = -3
NOTES
U1 +V3 – C13 = 0 + 8 – 12 = -4
U2 +V3 – C23 = 4 +7 – 12 = -2
U3 +V1 – C31 = 3 + 8 – 14 = -3
All Ui + Vj– Cij’s are negative so current basic solutions is the optimum
solution x11 = 900 ;
x21 = 300 ;
x22= 700 ;
x32 = 1000 ;
x33 = 200.
The minimum cost is 31,400 .
56 SRMIST DDE Self Instructional Material
2.6 ASSIGNMENT PROBLEM NOTES
Problem1.
Find the optimum assignments and processing cost for following problem.
Solution :
Step 1 - Row reduction
1st row minimum cost 19. Subtract 19 from other elements in 1st row
2nd row minimum cost 11. Subtract 11 from other elements in 2nd row
3rd row minimum cost 12. Subtract 12 from other elements in 3rd row
Step 2 - Column reduction
1st column minimum element is zero so no changes in 1st column
2nd column minimum element 3. Subtract 3 from other elements in 2nd
column
3rd column minimum element1. Subtract 1 from other elements in 3rd
column
57 SRMIST DDE Self Instructional Material
NOTES
Step 3 - Row scanning
If a row or column is having single zero allocate the zero.
Go to the first row. First row having single zero.
So allocate that zero and strike the first column.
No zero in second row so leave the second row go to third row.
And third row having two zeros, So do nothing.
Column scanning
First column already eliminated so go to the second column, Second column
contains single zero.
Allocate the zero and strike the entire row.
58 SRMIST DDE Self Instructional Material
NOTES
Third column contains no zeros. So row scanning and column scanning
process over.
The number of allocations should be equal to number of row or column.
But from the above table we observe,
number of allocations ≠ number of rows
So we should improve the current table to get optimum assignments.
Let = Minimum of undeleted elements in table
= Min{6,3,11,11}
=3
Subtract 3 from undeleted cell elements and add 3 in the cells where two lines
intersect in the above table.
Now again do row and column scanning, we get
59 SRMIST DDE Self Instructional Material
NOTES
Number of allocation = Number of rows, we reach the optimum solution .
Optimum Assignment: J1M1 ; J2M2; J3M3
Processing cost :19+17+15 = 51.
Problem 2.
Solve the following assignment problem and find the optimum assignment
and minimum processing cost.
Solution.
(The Number of rows )3 ≠ 4 (The number of columns ) ,
So the given problem is not balanced. So we include a dummy row with zero
cost to convert the unbalanced assignment problem in to balanced
assignment problem.
Coverted balanced assignment problem
60 SRMIST DDE Self Instructional Material
NOTES
The number of rows = The number of columns
So the above table represents a balanced assignment problem.
Find the minimum value in each row and subtract it from each element
in corresponding row.
Also find the minimum value in each column and subtract it from each
element in corresponding column.
Go to the first row, if the row having single zero then allocate it
otherwise do nothing, leave the row and go to the next row.
61 SRMIST DDE Self Instructional Material
After Row Scanning
NOTES
After column scanning
Number of allocations ≠ Number of rows. The allocation is not optimum.
Let =min{6,5,10,9,14,11,12}=5.
Add =5 at the cells, Where the line intersect and subtract =5 from the
undeleted elements in the above table.
62 SRMIST DDE Self Instructional Material
NOTES
Number of allocation ≠ Number of rows. Not a optimum solution.
Choose = Minimum of{5,4,6,7,9}=4 .
Subtract 4 from undeleted elements and add =4 at the cells where the lines
intersects
After row and column scanning we can reach two different optimum
solution.
63 SRMIST DDE Self Instructional Material
NOTES
Optimum assignment:
XA ; YB ; XC ; WD
Minimum cost 18+13+19+0 = 50
(OR)
Optimum assignments: XA ; YC ; XB ; WD
Minimum cost 1+17+15+0 = 50.
Exercise problems:
1. Find the initial basic feasible solution the following problem by
northwest corner rule.
64 SRMIST DDE Self Instructional Material
NOTES
Answer. x11=200 ; x12=50 ; x22=175 ; x23=125 ; x33=150 ; x34=250
Travelling cost is 12200.
2. Find the initial basic feasible solution by the transportation problem
by VAM method.
Answer. x11=40 ; x13=40 ; x14=120 ; x22=40 ; x23=140 ; x31=110 ; x44=50
Travelling cost is 13220.
Find the optimum basic feasible solution to the following transportation
problem by MODI method.
65 SRMIST DDE Self Instructional Material
1.
NOTES
Solution : x13=60 ; x21=50 ; x23=20 ; x32=80 travelling cost is 750.
2.
Solution : x13=2 ; x22=1 ; x23=2 ; x31=4 ; x33=1 travelling cost is 33.
3.
Answer. x11=3 ; x14= 2 ; x23=2 ; x32=3 ; x34=0; x24= 0
Transportation cost 32
Find the optimum assignments for the following assignment problems.
66 SRMIST DDE Self Instructional Material
NOTES
Answer. Optimum assignments: XA ; YB; ZC ;
Minimum cost is 50.
Answer. XA; YB; ZC; WD;
Minimumcost is 23.
Problem 3.
Answer. AG; BE; CF; DH.
67 SRMIST DDE Self Instructional Material
2.7 TRAVELLING SALESMAN PROBLEM
NOTES
Definition
The travelling salesman problem is to determine in what under the salesman
should travel, So that every city visited exactly once and finally return to the
starting point in such away that the total distance is minimum.
Problem 1.Solve the travelling salesman problem.
Solution.
Find the minimum value o each row and subtract it from each and every
element in the same row.
68 SRMIST DDE Self Instructional Material
Find the minimum value of each column and subtract it from every element
NOTES
in the corresponding column.
Go through the first row if the row is having single zero allocate the zero.If
the row having more than one zero move to the next row.
After row scanning go through the first column if the column is having only
one zero allocate the zero. If the column having more than one zero move to
the next column.
Number of allocations ≠ Number of rows.
So the allocation is not optimum. So improve the table and check the
optimality again.
Let = Minimum of {24, 49, 28, 3} = 3.
69 SRMIST DDE Self Instructional Material
Add the at the line of intersection and subtract from the undeleted
NOTES elements.
Optimum Assignment :A C B D A .
Travelling Cost :16 + 32 + 40 + 40 = 128.
70 SRMIST DDE Self Instructional Material
NOTES
Answer. ADBCEA
(or) AECBDA
Minimum cost 16 units.
Optimum assignments (i) AEBA and CDC (or)
(ii)ABEA and CDC
Minimum distance 363.
71 SRMIST DDE Self Instructional Material
2.8 TRANSHIPMENT PROBLEM
NOTES
A transportation problem in which available commodity frequently moves
from one source to another source or destination before reaching its actual
destination is called a transhipment problem.
2.8.1 Characteristics of transhipment problem
(i) The number of sources and destinations in the transportation
problem are m and n respectively. In transhipment problem,
however, we have m+n sources and destinations.
(ii) If Sidenotes the ith sources and Dj denotes the jth destinations then
commodity can move along the route SiDiDj, SiSjDiDj ,
SiDiSjDj or in various other ways. Clearly, transportation
cost from Si to Sj or Si to Di do not have to be symmetrical. In
general , SiSj≠ SiSj
(iii) In solving the transhipment problem, we first find the optimum
solutionto the transportation proble, and then proceed in the same
fashion as in solving the transportation problem.
(iv) The basic feasible solution contains 2m + 2n – 1 basic variables, if
we omit the variables appearing in the m+n diagonal cells, we are
left with m+n-1 diagonal cells, we are left with m+n-1 basic
variables.
Problem1. Solve the transhipment problem
72 SRMIST DDE Self Instructional Material
NOTES
Solution
Using VAM method and MODI method we get the optimum solution.
Optimum transhipment
Transport 25 units from S2 to the destination D1. It increases the
capacity of S2 to 55 units including 30 units as buffer stocks.
Transport 5 units from S1 to D2 .
Out of 25 units available at D1 transport 5 units from D1 to D
Transportation Cost: 25 × 2 + 5 × 4 + 5 × 1= 50 + 20 + 5 = 75.
73 SRMIST DDE Self Instructional Material
NOTES
CHAPTER III NETWORK ANALYSIS
74 SRMIST DDE Self Instructional Material
NOTES
STRUCTURE
3.1 Definition of Network link and path
3.2 Minimum spanning tree
3.3 Shortest path problems (using Djikstra’s algorithm)
3.4 Maximum Flow problem
3.5 Sequencing Problem
3.6 Decision Theory
3.1 DEFINITION OF NETWORK CONCEPTS
Definition - Network
A network of a set points and a set of lines connecting different pairs of
points. The points are called nodes and the lines are called links.
Definition - link
A link is said to be directed if it allows positive flows in one directions and
zero flow in opposite direction.
Definition - Path
A path is a sequence of distinct links that joint two nodes through other
nodes.
Definition - Cycle
A path is called cycle if it connects a node to itself through other nodes.
Definition - connected
A network is said to be connected if every two distinct nodes are connected
by at least one path.
Definition - tree
A tree is a connected network without any cycles.
75 SRMIST DDE Self Instructional Material
3.2 MINIMUM SPANNING TREE
NOTES
Problem 1: Determine under which roads telephone lines should be installed
to connect all towns with a minimum total length of lines. A cable company
providing cable services to fie new housing development areas the following
diagrams depicts the potential TV linkages among the five areas.
(a)Determines the most economical network for the company also find the
minimal spanning tree for the network.
(b)When nodes 2 and 5 cannot be linked.
(c) Nodes 5 and 6 linked by a 2 miles cable.
Solution.
Step 1:
Start from the node 1 the direct neighbours of 1 are 2,4 .among {2,4} 2 is the
nearest node 1 so connect 1 and 2.
76 SRMIST DDE Self Instructional Material
NOTES
Step 2:
Now identify the connected node that is nearest to one of the connected node.
The direct neighbours of 1 and 2 are {5,3,4} . Node 5 is nearest to 2.
Connect 2 and 5.
Step 3 :
The direct neighbours of the connected nodes 1,2,5 are 4 and 3.
Node 2 is nearest to4 so connect 2 and 4.
77 SRMIST DDE Self Instructional Material
NOTES
Step 4:
The direct neighbours of the connected nodes 1,2,4 ,5 are 3 and 6.
Node 4 and 6 is the nearest nodes so connect it.
Step 5:
3 is the only unconnected node nearest to 1 and 4. So connect 1 and 3 (or) 3
and 4.
78 SRMIST DDE Self Instructional Material
NOTES
a) The most economical network is 1-2-4-6,4-3,2-5
The required cable in miles 1+3+4+3+5=16 miles
b) Nodes 2 and 5 cannot be linked. In this case 5 is nearest to 4 so connect it
the required minimum spanning tree as follows.
Economical network is 1-2-4-6,4-3,4-5 ; The required cable in miles
1+4+3+5+8=21 miles.
(c) If nodes 5 and 6 are linked the minimal spanning tree is
79 SRMIST DDE Self Instructional Material
NOTES
Economical network in this case is 1-2-5-6-4-3 and required cable in miles
1++3+4+3+5=16 miles.
Problem 2
A company is in the process of providing pipeline network that
connects nine offshore natural gas wellheads with an inshore delivery
point. The location of wellhead 1 is the closest to the shore, so it is
equipped with adequate pumping and inventory capacity to pump the
output of the remaining eight wellheads to the delivery point. The
mileages of the feasible links (nodes 1 to 9) connecting to the inshore
delivery point are shown in the network below.
Solution.
Step1.
80 SRMIST DDE Self Instructional Material
We start at node 1. Connect node 5 to it being the nearest distinct node, NOTES
the link having the minimum length of 4.
The unconnected node closest to a connected node is node 6 (closest to
node 5). So we connect node 6 to node 5.
Now the unconnected node closest to a connected node is node 7.
So we connect node 7 to node 5.
Step 2.
The unselected node closest to a connected node is node 2. So we
connect node 2 to node 1.
Now the unticked node closest to a connected node is node 3. So we
connect node 3 to node 2.
Now the unticked node closest to a connected node is node 9. So we
connect node 9 to node 5.
81 SRMIST DDE Self Instructional Material
NOTES
Step 3.
The unselected node closest to a connected node is node 8. So we
connect node 8 to node 9.
Now only unselected node is node 4 with the closest connected node
being node 6. So, we connect node 4 to node 6.
82 SRMIST DDE Self Instructional Material
Final step NOTES
Now all the nodes have need connected. The minimal spanning tree
connecting all nodes is displayed by links shown thick in the network
above.
Thus the most economical pipeline network for the company that links
the offshore wellheads to the inshore delivery point is given by the
nodes connected in the following order.
Economical path:
1 – 5 – 6 ; 5 – 7; 1 – 2 – 3 ; 5 – 9 – 8 ; 4-6
The optimal length for this minimal spanning tree is given by
Z = 4 + 3 + 5 + 5 + 6 + 6 + 5 + 7 = 41 miles.
83 SRMIST DDE Self Instructional Material
NOTES
3.3 SHORTEST PATH PROBLEMS (USING DJIKSTRA’S
ALGORITHM)
Problem1.
Determine the shortest path from A to C for the following network.
Solution.
Step 1.
Label the starting node A as (0, - ) as it is the origin.
As no further improvement is feasible for this node, we this label as final.
Select this node to indicate that this node has been solved.
Step 2.
84 SRMIST DDE Self Instructional Material
Nodes directly connected to the last ticked node are B, E and D. We compute
NOTES
the working labels for each one of them
B = (1, A ), E = (4, A) and D = (5, A).
Step 3. Choose the unselected node with the smallest working label, having
smallest ui , and record its working label as its final label. So we tick node B to
indicate that this node has been solved.
Step 4.
85 SRMIST DDE Self Instructional Material
Un selected nodes directly connected to the last ticked node are nodes E and
NOTES C.
Compute or update the working labels for nodes E and C .
Step 5.
Choose the unselected node with smallest working label and mark its label as
final. Select node E to indicate that the node has been solved.
Step 6.
Update working lables of all unticked nodes.
86 SRMIST DDE Self Instructional Material
Choose the unselected node with smallest working label and mark its label as
NOTES
final.
So we tick node D indicate that this node has been solved.
Step 7.
The unselected node directly connected to the last ticked node D is node C.
Thus node C gets another working label (7, D).
However the new label gives no improvement and is therefore as good as the
previous label.
Moreover as node C happens to be the destination node, it does not lead to
any further node.
So we make its both the working labels final and select the node C to indicate
this node having been solved.
Step 8.
All nodes from origin to destination have been ticked. So we can stop.
Step 9.
The shortest path from node A to C is found by back tracking as follows.
( C ) ( 7 , B ) ( B ) ( 1 , A ) ( A )
Thus the shortest path is
A BC and the length of the path is z = 7.
87 SRMIST DDE Self Instructional Material
(OR)
NOTES
( C ) ( 7 , D ) ( D) ( 5, A ) ( A )
A DC and the length of the path is z = 7.
Problem 2.
Determine the shortest route from A to E in the following network.
Solution.
Step 1:
Select the starting node and label A as (0, -).
Step 2:
The unselected nodes directly connected to A are B and C .
The distances from A to B and C are 4,3 respectively.
So, allocate the label B=(4,A)and C=(3,A).
88 SRMIST DDE Self Instructional Material
C has the minimum distance from A so select C.
NOTES
Step 3:
The unselected nodes directly connected to C is D.
The distance from C to D is 4 and C already has working label (3, A)
3+4=7, so allocate the label D=(7,C).
Now B=(4,A) & D=(7,C),among B,D the node B having the smallest working
label 4.
Select the node B.
Step 4:
The unselected node directly connected to B is D now D gets new label (12,D)
89 SRMIST DDE Self Instructional Material
However the new label gives no improvement so keep the old label (7, C).
NOTES select the Node D.
Step 5:
The unselected node directly connected to Dis E.
The distance between E-D is 7 and D has a working label (7,C) so 7+7=14 and
E=(14,D) select the node E
All the nodes from origin to destination have been selected so stop the
algorithm.
The shortest path from A to E is
A-C-D-E
Minimum distance is14.
90 SRMIST DDE Self Instructional Material
NOTES
3.4 MAXIMUM FLOW PROBLEM
Procedure
Find a path from Soure to Sink.
Identify the bottle neck capacity or flow capacity of the path
k = min of capacity of each link in the path
Subtract k from the capacity of each link in the selected path.
If the capacity of a link becomes zero block the path.
Repeat the process until all the path from Source to Sink are blocked.
Problem 1.
For the network shown, below find the flow pattern giving the maximum
flow from the source to the sink, given that the link capacity from node i to
the node j is the number nearest node i along the link between these nodes.
Node S represents the source and T represents the sink in the following
diagram.
Solution.
Step 1.
Find a path from S – T and identify its bottle neck capacity or flow capacity
Path: S – A – B – T
Flow = min{ 10, 4, 10} = 4.
Subtract the flow 4 from capacity of each link in the path.
91 SRMIST DDE Self Instructional Material
The link A – B reaches its full capacity so the link A – B is blocked.
NOTES
The total output at the sink T is 4
Step 2.
The link A – B is blocked so, we find another path from S to T via unblocked
links.
Path: S – C – D – T
Flow = min{ 10, 9, 10} = 9.
Subtract the flow 9 from capacity of each link in the path.
The link C – D reaches its full capacity so the link C – D is blocked.
Now, the total output at the sink T is 4+9.
92 SRMIST DDE Self Instructional Material
Step 3.
NOTES
The link C – D and A – B are blocked so, we find another path from S to T via
unblocked links.
Path: S – A – D – T
Flow = min{ 6, 8, 1} = 1.
Subtract the flow 1 from capacity of each link in the path.
The link D – T reaches its full capacity so the link D – T is blocked.
Now, the total output at the sink T is 4+9+1.
Step 4.
The links D – T, C – D and A – B are blocked so, we find another path from S
to T via unblocked links.
Path: S – A – D – B - T
Flow = min{ 5, 7, 6, 6} = 5.
Subtract the flow 5 from capacity of each link in the path.
The link S – A reaches its full capacity so the link S – A is blocked.
Now, the total output at the sink T is 4+9+1+5.
93 SRMIST DDE Self Instructional Material
NOTES
No other route is possible via unblocked links from the Source S to the
sink T. So the maximal flow is 19.
Exercise problem
Problem 1.
Consider the following network where the numbers on links represent actual
distance between the corresponding nodes. Find the minimal spanning tree:
Answer. O – A – B – C ; B – E – D –T ; Minimum length z = 18.
Problem 2.
The management of City Information Park desires to determine which road
telephone cables should be installed to connect all 7 work stations with a
minimum total length of cable. Nodes and distances of potential links are
given in the network below. Use the minimal spanning tree algorithm to find
the most economical cable tree.
94 SRMIST DDE Self Instructional Material
NOTES
Answer: . O – A – B – C ; B – E – D –T ; Minimum length z = 14.
Problem 3
The following network gives the links and associated length in miles between
city1 and four other cities(nodes 2 to 5). Use Dijkstra’s labeling procedure to
determine the shortest route between city 1 and city2.
Answer: 1 – 3 – 4 – 2 ; Shortest route 55 miles.
Problem 4.
For the network shown, below find the flow pattern giving the maximum
flow from the source to the sink, given that the link capacity from node i to
the node j is the number nearest node i along the link between these nodes.
Node O represents the source and S represents the sink.
95 SRMIST DDE Self Instructional Material
3.5 SEQUENCING PROBLEM
NOTES
Definition
It is a technique used to select the best sequence of processing different jobs
from different machines such that the total processing time of jobs and the
idle time of the machines should be minimum.
Sequencing problems
N jobs 2 machine optimum sequence algorithm
Step 1:Find the minimum of {A1,A2….An, B1, B2….Bn}.
Step 2:If the minimum be Ak, do the kth job first.
Step 3: If the minimum be Br, do the rth job last.
Step 4:
If the ties occurs (ie) Ak=Br for some k, r. Process the kth job first rth job
last.
If tie occurs among Ai’s then compare the corresponding Bi’s value.
Process the job corresponding to the maximum value of Bi first and
minimum value of Bi second.
If the tie occurs among Bi’s then compare the corresponding value of
Ai.
Process the job corresponding to the maximum Ai last. The job
corresponding to the minimum Ai processed next to the last.
Step 5: Eliminate the job assigned and repeat the process until all the job have
been assigned.
96 SRMIST DDE Self Instructional Material
PROBLEM 1:
NOTES
Determine the optimum sequence for performing the job that would
minimize the total elapsed time.
Solution.
The Smallest processing time in given problem is 1,on machine A.
So perform the corresponding job J1 first
The Minimum processing time in the above table is 2. Which corresponds to
J4 and J5 .
Tie occurs in Machine B then compare the corresponding processing times in
Machine A the minimum of {5,6} is 5 which is corresponds to J4.
So process the J5 last and J4 shall be processed next to the last.
97 SRMIST DDE Self Instructional Material
NOTES
There is tie among 3 Jobs for the smallest processing time in this reduced
problem.
These corresponds to J2 and J6 on Machine A and to J3 on Machine B. As the
corresponding processing time of J6 on Machine B is larger than the
corresponding processing time of J2 on Machine B. J6 will be processed next
to J1. Process J2 after J6 the updated job sequence is
J1 J6 J2 J3 J4 J5
Idle Time for Machine A is 3.
98 SRMIST DDE Self Instructional Material
Idle Time for Machine B is 1.
NOTES
Total Minimized processing time is 29 hours.
Problem 2.
A book binder has one printing press, one binding machine and the
manuscripts of a number of different books. The time required to perform the
printing and binding operations for each book is shown below. Determine the
order in which books should be processed in order to minimize the total time
required to tur out all the books.
Solution.
Her the books will first go to the printing press and the on the binding
machine.
If Pi denotes the time in hours on printing press and
Bi denotes the time in hours on Binding time.
Minimum{ Pi , Bi } = 10 corresponding to B6 book will be processed in the last.
The job sequence will be
The problem then reduces to the following five jobs two machines.
Now Minimum{ Pi , Bi } = 20 which corresponds to the P4.
Therefore, book 4 will be processed just in the beginning.
The entries are shown in the sequence cells as below.
99 SRMIST DDE Self Instructional Material
NOTES
After assigning books 4 and 6 , we now left with the four books and two
machines.
The reduced processing times as follows.
Now Minimum{ Pi , Bi } = 30 which corresponds to the P1 and B5.
There is a tie for the minima.
Therefore, book 1 will be processed just in the beginning.
And the book 5 will be processed next to the last
The entries are shown in the sequence cells as below.
We now left with the problem of 2 jobs and 2 machines with their respective
processing times as follows
The smallest printing time is 50 for book 3, we place book 3 in the third cell
and the remaining book 2 in the fourth cell and get the optimal sequence.
100 SRMIST DDE Self Instructional Material
The minimum elapsed time from the start of the first book to the completion
NOTES
of the last book corresponding the optimal sequence is computed as shown in
the following table.
From the above table it is clear that minimum elapsed time is 430 hours.
Idle time for the printing machine is 10 hours.
Idle time for the binding machine is 60 hours.
101 SRMIST DDE Self Instructional Material
Step 1:
NOTES
Find the Minimum of {A1 ,……An}, Minimum of {C1,C2……Cn} and
Maximum 0f {B1,B2,…..Bn}.
Step 2:
Check the conditions
1. Min {Ai’s} ≥ Max {Bi’s}
2. Min {Ci’s} ≥ Max {Bi’s}
Step 3:
If at least any one of the above two conditions is true then we can convert the
3 Machine problem in to 2 Machine problem.
Step 4:
Convert the 3 Machine Jobs by 2 Machine Jobs by introducing two imaginary
Machines X , Y such that
Step 5:
Determine the optimum sequence for n Jobs and 2 Machines. The resulting
sequence shall be the Optimumjob sequence for the given problem.
Note:
The above mentioned procedure is suitable for the problems satisfying any
one of the conditions mentioned in step2 . There are many practical problems
does not obey this rule.
102 SRMIST DDE Self Instructional Material
Problem 2 : Determine the optimum sequence of Jobs that minimizes the total
NOTES
elapsed time based on the following information processing time is given in
hours and passing is not allowed .
Solution :
The Minimum processing time of M1 is 3.
The Minimum processing time of M3 is 5.
The Maximum processing time of M2 is 5 .
One of the condition is true so we can convert the 3 Machine problem into 2
Machine problem.
103 SRMIST DDE Self Instructional Material
NOTES
The next Minimum processing time is 10. Tie occurs in three places. It
corresponds to G on Machine X.
B, F on Machine Y. The corresponding processing time of B on X is which is
lesser than the processing time F on X.
Therefore process F last and B next to the last.
104 SRMIST DDE Self Instructional Material
Total elapsed Time: NOTES
Total processing time is 59 hours
Idle time for M1 = 13 hours
Idle time for M2 = 37 hours
Idle time for M3 = 7 hours
Exercise problems
Problem 1.
We have five jobs, each of which must go through the two machines A and B
in the order AB. Determine a sequence for the five jobs that will minimize the
elapsed time. Processing times in hours are given in the table below:
Answer. Job sequence 2 – 4 – 3 – 5 -1 ; Total time 30 hours.
105 SRMIST DDE Self Instructional Material
Exercise 2.
NOTES
A ready-made garment company has to process 7 times through two stages of
production, cutting and sewing. Find an order in which these seven items are
to be processed so as to minimize the total processing time. The time taken by
each of these items at the different stages are gives as:
Answer. 3 – 5 – 7 – 6 – 2 – 4 – 1 ; Minimum processing time 47.
Exercise 3.
Determine the optimum sequence of Jobs that minimizes the total elapsed
time based on the following information processing time is given in hours
and passing is not allowed .
Answer. 1 – 3 – 2 – 4 – 6 – 5 ; Minimum processing time 67.
Exercise 4.
We have 7 jobs, each of which must go through machines A, B and C inorder
to A B C. Processing times are given in the following table: Determine the
optimal sequence and total elapsed time associated with it.
106 SRMIST DDE Self Instructional Material
Answer. 7 – 2 – 1 – 3 – 5 – 6 - 7; Minimum processing time 102hrs.
NOTES
Idle time A-16 hrs B-80 hrs and C-17 hrs.
107 SRMIST DDE Self Instructional Material
3.6 DECISION THEORY
NOTES
Definition
Decision theory is an integral part of management, planning, organizing and
controlling process. The decision maker selects one strategy over others
depending on some criteria.
Decision making process
The decision making process involves following major steps:
Determine various alternative courses of action from which the final
decision is to be made.
Identify the possible outcomes, called the states-of-nature or event for
the decision problem. The events are beyond the control of the
decision-maker.
Determine the payoff function which describes the consequences
resulting from the different combinations of the acts and events. The
payoffs may be designated as aij’ s the payoff resulting from the
combination of ith event and the j th action.
Construct the regret or opportunity loss table. An opportunity loss
occurs due to failure of not adopting the best available course of
action. The opportunity loss values are calculated separately for each
state-of-nature by first locating the most favorable course of action for
the payoff value for the best possible course of action that could have
been selected.
There are some certain types of environment according to which decision are
make. The types of environment are
Decision under certainty
Decision under conflict
Decision under uncertainty
Decision under risks
108 SRMIST DDE Self Instructional Material
Decision under certainty NOTES
Whenever there exists only one outcome for a decision, we are dealing with
this category.
Example: Linear programming, transportation, assignment etc.
Decision under conflict
In many situations, neither states-of-nature are completely known nor are
they completely uncertain. Partial knowledge is available and therefore it
may be termed as decision making under partial uncertainty.
Example: The situation of conflict involving two or more competitors
marketing the same product.
Decision under uncertainty.
These refers to situations where more than one outcome can result from any
single decision. These shall now we discussed.
Decision under risk
These refer to decision situations wherein the decision maker chooses from
among several possible outcomes where probabilities of occurrence can be
stated objectively from the past data.
Illustration:
First note that the short-sighted approach of selection the cheapest run
offered by each successive stage need not yield an overall optimum decision.
Following this strategy would give the route ABFIJ, at a total cost of
13.
However,sacrifice a little on one stage may permit greater savings thereafter.
Forexample,ADF is cheaper overall than ABF.
One possible approach to solve this problem is to use trial and error.However
the number of possible routes is large (18),and having to calculate the total
cost for each route is not an appealing task.
Figure:
109 SRMIST DDE Self Instructional Material
The road system and costs for the stagecoach problem.
NOTES
Formulation:
Let the decision variables xn (n=1, 2, 3, 4) be the immediate destination on
stage n (the nth stagecoach run to be taken).
Thus the route selected is A->x1->x2->x3->x4 ,where x4=J .
Let fn (s,xn) be the total cost of the best overall policy for the remaining stage
,given that the fortune seeker is in state S,ready to start stage n ,and select xn
as the immediate destination .
Given s and n, letxn* denote any value of xn (not necessary unique) that
minimizes fn(S,xn) and let fn*(s) be the corresponding minimum value of
fn(S,xn) .
Thus ,
Fn*(s)=min fn(s,xn)=fn(s,xn*)
Where fn(s,xn) isgiven by the preceding tables for cij by setting i=s(the
current state)and j=x (the immediate destination) .
Because the ultimate destination (state J) is reached at end of stage 4,
f5*9(J)=0
110 SRMIST DDE Self Instructional Material
The objective is to find fi*(a) and the corresponding route.
NOTES
Dynamic programming finds it by successively finding f4*(s), f3*(s), f2*(s),for
each of the possible states s and then using f2*(s) to solve for f1*(A).
Solution procedure:
When the fortune seeker has only one more stage to go (n=4), his route
thereafter is determined entirely by his current state S (either H or I) afinal
destination x4=J.
So the route for this final stagecoach run is S>J.
Therefore, since
f4*(s)=f4 (S.J)=cSJ the immediate solution to the n=4 problem is n=4:
When the fortune seeker has two more stage to go (n=3),the solution
procedure requires a few calculations.
For example ,suppose that the fortune seeker is in state F. then,as depicted
below, he must next go to either state H or Iat an immediate cost of Cfh=6 or
Cfi=3 ,respectively .
If he chooses state H, the minimum additional cost after he reaches there is
given in the preceding table as f4*(H) =3 ,as shown above the H node in the
diagram .
Therefore, the total cost for this decisions is 6+3=9.
If he chooses state I instead, the total cost is 3+4=7, which is smaller.
Therefore, the optimum choice is the latter one, x3*=I, because gives the
minimum cost f3*(F) =7.
111 SRMIST DDE Self Instructional Material
NOTES
Similar calculations need to be made when you start from the other two
possible states s=E and s=G with two stages to go .Try it, proceeding both
graphically and algebraically [combining cij and f4*(s) values], to verify the
following complete results for the n=3 problem .
n=3:
The solution of the second–stage problem (n=2), where there are three stages
to go, is obtained in a similar fashion .
In this case, f2(s, x3) =csx2+f3*(x2).for example, suppose that the fortune
seeker is in state C,as depicted below.
112 SRMIST DDE Self Instructional Material
NOTES
He must go to state E,F or G at an immediate cost of CCE=3 ,CCF=2, or
CCG=4 ,respectively .
After the getting there ,the minimum additional cost for stage 3 to the end is
given by the n =3 table as f3*(E)=4,f3*(F)=7,or f3*(G)=6 ,respectively ,as
shown above the E and F nodes and below the G node in the preceding
diagram .
The resulting calculations for the three alternatives are summarized below.
X2=E: f2(C, E) =CCE+f3*(E) =3+4=7.
X2=F: f2(C, F) =CCF+f3*(F) =2+7=9.
X2=G: f2(C, G) =CCG+f3*(G) =4+6=10.
113 SRMIST DDE Self Instructional Material
The minimum of these three numbers is 7.
NOTES
So the minimum total cost from state cost to the end is f2*(C)=7, and the
immediate destinations should be x2*=E.
Making similar calculations when you start from state B or D (try it) yields
the following results for the n=2 problems:
In the first and third rows of this table,note that E and F tie as the minimizing
value of x2.
so the immediate destinations from either state B or D should be x2*=E or F .
Moving to the first-stage problem (n=1), with all four stages to go, we see that
the calculations are similar to these just shown for the second-stage problem
(n=2) ,except now there is just one possible starting s=A ,as depicted below .
114 SRMIST DDE Self Instructional Material
These calculations are summarized next for three alternatives for the
NOTES
immediate destination:
x1=B: f1(A,B)= CAB+f2*(B)=2+11=13 .
x1=C: f1(A,C)= CAC+f2*(C)=4+7=11 .
x1=D: f1(A,D)= CAB+f2*(D)=3+9=11 .
In the above Graphical display of the dynamic programming solution of the
stage coach problem,each arrow shows an optimal policy decision(the best
immediate destination) from that state, where the number by the state is the
resulting cost from there to the end.
Following the bold face arrows from A to T gives the three optimal solutions(
the three routes giving the minimum total cost of 11).
Since 11 is minimum. f1*(A)=11 and x1*=C and D,as shown in the following
table .
115 SRMIST DDE Self Instructional Material
NOTES
An optimal solutions for the entire problem can now be identified from the
four tables .Results for the n=1 problems indicate that the fortune seeker
should go initially to either state C or state D .
Suppose that he chooses x1*=C .for n=2 the results for s =C is x2*=E .this
results leads to the n=3 problems, which gives x3*=H for s=E ,and the n=4
problem yields x4*=J for s=H
Hence, one optimal route is ACEHJ .
Choosing x1*=D leads to the other two optimal routes
ADEHJ and ADFIJ .
They all yield a total cost of f1*(A)=11 .
116 SRMIST DDE Self Instructional Material
3.7 REPLACEMENT PROBLEMS NOTES
Definition:
A replacement problem is to determines the optimum replacement age of an
equipment/object where maintenance cost increases with time and value of
the money remains static during that period .
Notation and formulae:
C-capitals cost of equipment
S-scrap value
n-no of years
f(n)-maintenance cost function at the nth year
A(n)-average total annual cost
S(n) – Resale value at the end of nth year.
D(n)=S(n-1)-Sn
E(n)=d(n)+f(n)
Problem: 1
A firm is considering replacements of machine , where the cost price is
RS.12,200 and the scrap value is RS.200 the running cost in rupees are found
from experience be as follows
Year 1 2 3 4 5 6 7 8
Running 200 500 800 1200 1800 2500 3200 4000
cost
When should the machine replaced ?
117 SRMIST DDE Self Instructional Material
NOTES
From the above table we observe the average cost in the 6 year is minimum .
Also the average in the 7 years the year is more than the cost in the 6 th year
hence the machine should be replaced after 6 years.
Replacement policy when value of money changes with time
Rn - operating cost in the year n
V - discount rate (or) present worth factor
If money carry a rate of interest per year then
V=(1 + 𝑟)−1
Problem 1.
The cost of a new machine is RS.5000 the maintenance cost of nth year is
given by 500(n-1),n=1,n=2……
Suppose that the discount rate per year is 0.5.after how many years it will be
economical to replace the machine by new one?
118 SRMIST DDE Self Instructional Material
Solution: NOTES
The discount rate of money per year is 0.05 , the present worth factor
V=(1 + 𝑟)−1 =(1 + 0.05)−1 =0.9523
The optimum replacement time is determined in the following table
Since ,W(n) is the minimum for n=5 and 𝑅4 (=1500)<W(5) as well as
W(5)>𝑅6 (=2500) it is economical to replace the machine by a new one at the
end of five years.
Problem 2:
A manufacturer is offered two machine A and B. Machine A is priced at
Rs.5000,and running costs are estimated at Rs.800 for each of the first five
years ,increasing by RS.200 per year in the sixth and subsequently years .
Machine B, which has the same capacity as A, costs Rs.2500 but will have
running costs of Rs.1200 per year, for 6 years, increasing by the Rs. 200 per
year thereafter if the money is worth 10 percent per year, which machine
should be purchased? (assume that the machine will eventually be sold for
scrap at a negligible price)
119 SRMIST DDE Self Instructional Material
Solution:
NOTES
Since the money Is worth 10% per year , the discount rate for both the
machine is given by
1
V = 1+0.10 = 0.9091
For the solution of this problem, we compute the following tables for
machines A and B separately ,by using Pwf table given at the end of the book.
120 SRMIST DDE Self Instructional Material
NOTES
From the above table we observe that for machine A,
1600 <1751.72 < 1800.
Now , since the running cost of 9th year is Rs.1600 and that of 10th is Rs.1800
and since 1800 > 1751.72, it is better to replace the machine A after 9th year.
Similarly ,for machine B since 1800>1680.23,it is better to replace the machine
B after 8th year.
Further since the weighted average cost in 9 years of machine A Rs.1751.72
and the weighted average cost in 8 years of machine B is Rs.1680.23
It is advisable to purchase machine B.
121 SRMIST DDE Self Instructional Material
Replacement of equipment that fails suddenly
NOTES
Under this circumstance we consider two types of replacements policies.
Individual replacements policies: under this policy, an item is replaced
immediately after its failure.
Group replacements policy:under this policy, we take decisions as when all
the items must be replaced ,irrespective of the fact that items have failed or
have not failed, irrespective with a provision that if any item fails before the
optimal time it may be individually replaced.
N =total number of items in the system,
C2 =cost of replacing an individual item,
C1 =cost of replacing an item in group,
C(t) =total cost of group replacements after time period t,
F(t) =number of failures during time period t
Problem1.
The following failure rates have been observed for a certain type of
transistors in a digital computer:
The cost of replacing an individual failed transistor is Rs.1.25. the decision is
made to replace all these transistors simultaneously at fixed intervals, and to
replace the individual’s transistors as they are in service.
If the cost of group replacement 30 paise per transistor, what is the best
interval between group?
At what replacement price per transistor would a policy of strictly individual
become preferable to the adopted policy?
122 SRMIST DDE Self Instructional Material
Solution: NOTES
Suppose there are 1000 transistors in use. Let Pi be the probability that a
transistor, which was new placed in position for use, fails during the ith week
of its life.
Thus, we have
P1 =0.05
P2 =0.13-0.5 =0.08
P3 =0.25-0.13 =0.12
P4 =0.43-0.25 =0.18
P5 =0.68-0.43 =0.25
P6 =0.88-0.68 =0.20
P7 =0.96-0.88 =0.08
P8 =1.00-0.96 =0.04
Let Ni denote the number of replacements made at the end of the ith week.
Then, we have
𝑁0 = number of transistors in the beginning =1000
𝑁1 =𝑁0 𝑃1 = 50
𝑁2 =𝑁0 𝑃2 +𝑁1 𝑃1 = 82
𝑁3 =𝑁0 𝑃3 +𝑁1 𝑃2 +𝑁2 𝑃1 = 128
𝑁4 =𝑁0 𝑃4 +𝑁1 𝑃3 +𝑁2 𝑃2 +𝑁3 𝑃1 =199
𝑁5 =𝑁0 𝑃5 +𝑁1 𝑃4 +𝑁2 𝑃3 +𝑁3 𝑃2 +𝑁4 𝑃1 =289
𝑁6 =𝑁0 𝑃6 + 𝑁1 𝑃5 + 𝑁2 𝑃4 + 𝑁3 𝑃3 + 𝑁4 𝑃2 + 𝑁5 𝑃1 =272
𝑁7 =𝑁0 𝑃7 + 𝑁1 𝑃6 + 𝑁2 𝑃5 + 𝑁3 𝑃4 + 𝑁4 𝑃3 + 𝑁5 𝑃2 + 𝑁6 𝑃1 = 194
𝑁8 =𝑁0 𝑃8 + 𝑁1 𝑃7 + 𝑁2 𝑃6 + 𝑁3 𝑃5 + 𝑁4 𝑃4 + 𝑁5 𝑃3 + 𝑁6 𝑃2 +𝑁7 𝑃1 = 195
From the above calculation, we observe that the expected number of
transistors failing each week increases till 5th week and then starts decreases
and later again increasing from 8th week.
123 SRMIST DDE Self Instructional Material
Thus Ni will oscillate till the system acquires a steady state. The expected life
NOTES of each transistor is
= 1×0.5+2×0.08+3×0.12+4×0.18+5×0.25+6×0.2+7×0.08+8×0.04
= 4.62 weeks.
Average number of failures per week
= 1000/4.62 = 216 approximately
Therefore the cost of individual replacements
= 216×1.25 = Rs.270.00 per week.
Now, since the replacement of all the 1000 transistors simultaneously cost 30
paise per transistors and the replacements of an individual transistor an
failure cost rs.1.25, the average cost for different group replacements policies
is given as under:
Since the average cost is lowest week 3, the optimum interval between group
replacements is 3weeks.
Further, since the average cost is less than Rs.270 (for individual replacements
the policy of group replacements is better.
124 SRMIST DDE Self Instructional Material
Problem 2. NOTES
A computer has a large number of electronics tubes .they are subject to
mortality as given below:
If the tubers are group replaced, the cost of replacements is Rs.15 per tube.
Group replacement can be done at fixed interval in the night shift when the
computer is not normally used of individual tubes which fail in service costs
Rs.60 per tube .
How frequently should the tubes be replaced?
Solution:
Consider each block of 200 hours as one period and assume that there are
1000 tubes initially.
Let Ni be the numbers of replacements made at the end of the ith period, if all
the 1000 tubes are new initially .then the expected numbers of failures at
different weeks can be calculated as shown below:
𝑁1 =𝑁0 𝑃1 = 1000×0.10=100
𝑁2 =𝑁0 𝑃2 +𝑁1 𝑃1 =1000×0.26+100×0.10=270
𝑁3 =𝑁0 𝑃3 +𝑁1 𝑃2 +𝑁2 𝑃1 =1000×0.35+100×0.26+270×0.10=403
𝑁4 =𝑁0 𝑃4 +𝑁1 𝑃3 +𝑁2 𝑃2 +𝑁3 𝑃1 =1000×0.22+100×0.35+270×0.26+403×0.10=365
𝑁5 =𝑁0 𝑃5 +𝑁1 𝑃4 +𝑁2 𝑃3 +𝑁3 𝑃2 +𝑁4 𝑃1 =1000×0.07+100×0.22+270×0.35+403×0.26+36
5×0.10=328
125 SRMIST DDE Self Instructional Material
From the above calculation, we observe that the number of tube of tubes
NOTES failing in each period increases till the third period and then starts decreasing.
Thus the value of Ni will oscillate till the system settles down to a steady
state.
In the steady state, the proportion of tubes failing during each period is the
reciprocal of their average life.
Expected life of a tube
=1×0.10+2×0.26+3×0.35+4×0.22+5×0.07=2.90 periods
Expected numbers of failures per period
= 1000/2.90=345
Cost of individual replacements per period
= 345×60 = Rs.20700
Now, since the replacements of all the 1000 tubes simultaneously cost Rs.15
per tube and the replacements of an individual tube on failure cost Rs.60 the
average cost for different group replacements policies is given as follows:
Since the average cost is lowest against period 2, the optimum interval group
replacements is two periods (i.e)
After 400 hours further since the average cost of group replacements (which
is Rs.18600 is less Rs.20700 (cost of individual replacements) the policy of
group replacement is better.
126 SRMIST DDE Self Instructional Material
Exercise 1.
NOTES
The cost of a machine is Rs 6100 and its scrap value is Rs. 100. Find when the
machine be replaced? The maintenance costs found from experience are as
follows:
Exercise 2.
A firm considering replacement of a machine whose cost price is Rs. 17500
and the scrap value is Rs. 500. The maintenance costs are found from
experience to be as follows. Find when the machine should be replaced?
Exercise 3.
In a machine shop, a particular cutting tool cost Rs 6 to replace. If the tool
breaks on the job, the production disruption and associate costs amount Rs
30. The past life of tool is given as follows
After how many jobs, should the shop replace a tool before it breaks down?
Problem 4.
The failure rates have been observed for a certain type of bulb:
127 SRMIST DDE Self Instructional Material
NOTES
There are 1000 bulbs in use. The cost of replacing an individual failed bulb is
Rs. 1.25. If all the bulbs are replaced as a group, it costs Rs 0.60 per bulb.
Determine among individual and group replacement policies which one is
better.
Problem 5.
The following failure rates have been observed for each certain type of light
bulbs
week 1 2 3 4 5
Percent of 10 25 50 80 100
failure
There are 1000 bulbs in use, and it costs Rs 2 to replace an individual bulb
which has burnt out.
If all bulbs were replaced simultaneously, it would cost 50 paise per bulb.
It is proposed to replace all bulbs at fixed intervals, whether or they have
burnt out, and to continue replacing burnt out bulbs as they fail.
At what interval should all the bulbs be replaced?
128 SRMIST DDE Self Instructional Material
CHAPTER IV QUEUEING THEORY NOTES
STRUCTURE
4.0 Queueing Theory concepts
4.1 Model I - (M/M/1) : (∞/FIFO)
4.2 Model II - (M/M/1) : (𝑵/FIFO)
4.0 QUEUEING THEORY CONCEPTS
Queue
A flow of customers from infinite/finite population towards the service
facility forms a queue on account of lack of capability to serve them all at a
time.
Customer
By the term customer we mean the arriving unit that requires some service to
be performed. The customers may be of persons, machines, etc.
Service channel
The process that performs the services to the customer is termed by service
channel
Pattern of arrivals:
customers may arrive in the system at known times in a random way. This
type of queuing problems are categorized as deterministic models.
On the other hand, if the time between successive arrivals is uncertain, the
arrival pattern is measured by mean arrival rate.
These queues are characterized by the probability distribution associated
with this random process.
Queue discipline:
129 SRMIST DDE Self Instructional Material
It is a rule according to which customers are selected for service when a
NOTES queue has been formed. The most common queue discipline is the first come
first out.
Another queue discipline is last in first out. This queue discipline practised in
most cargo handling situations where the last item loaded is removed first.
Service mechanism:
The service mechanism is concerned with service time and service facilities.
Service time is the time interval from the commencement of service to the
completion of service .
Capacity of the system:
The source from which customers are generated may be finite or infinite. A
finite source limits the customers arriving for service.
Transient state:
A queuing system is said to be in transient state when its operating
characteristic mean queue length are dependent upon time.
Steady state:
If the characteristic of the queuing system becomes independent of time, then
the steady-state condition is said to prevail.
Classification of queuing Models
Generally queuing models may be completely specified in the following
symbolic form (a/b/c) : (d/e) .
The first and second symbols denote the type of distributions of inter-arrival
times and of inter-service times, respectively.
Third symbol specifies the number of servers, whereas fourth symbol stands
for the capacity of the system and the last symbol denotes the queue
discipline.
130 SRMIST DDE Self Instructional Material
4.1 MODEL I - (M/M/1) : (∞/FIFO) NOTES
To find the average number of units in the system,
𝛒 𝛌 𝝀
Ls = = where ρ = µ and ρ <1 .
(𝟏− 𝛒) (µ− 𝛌)
To find the average length of queue (or) the number of customers in the
queue, excluding the customer which is in service
𝛒𝟐 𝛌𝟐
L q= Ls – ρ = = .
(𝟏− 𝛒) µ(µ− 𝛌)
Expected waiting time in the system
𝑳𝒔 𝟏
Ws = = .
𝛌 µ −𝛌
Waiting time in the queue
𝑳𝒒 𝛌
Wq= = .
𝛌 µ(µ − 𝛌)
Expected waiting time of a customer
𝟏
(µ − 𝛌)
.
Expected length of a non empty queue
µ
.
(µ − 𝛌)
Probability of queue size ≥ N is ρN.
Probability of waiting time in a system is
∞
𝒕
(µ − 𝛌)𝒆−(µ − 𝛌)𝒘 𝒅𝒘.
Probability of waiting time in the queue
∞
𝒕
𝛒 (µ − 𝛌)𝒆−(µ − 𝛌)𝒘 𝒅𝒘.
𝛌
Traffic indensity .
µ
Problem 1.
131 SRMIST DDE Self Instructional Material
In a railway Marshalling yard, goods train arrive at a rate of 30 trains per day.
NOTES Assuming that inter arrival time follows an exponential distribution and the
service time distribution is also exponential, with an average of 36 minutes.
Calculate the following:
The mean Queue size(line length)
The probability that queue size exceeds 10
If the input of the train increases to an average 33 per day what will be the
changes in (i) and (ii)?
Solution:
30 1
λ= = and
60X24 48
1
µ= trains perminute
36
We know that
𝜆 36
ρ = µ = 48 = 0.75
ρ 0.75
L s= = = 3 trains.
(1− ρ) 1−0.75
P(≥ 10) = 0.7510 = 0.056 .
When the input increases to 33 trains per day then we get
33 1
λ= = ,
60X24 480
1
µ= trainsperminute
36
𝜆
andρ = = 0.825.
µ
Now
ρ 0.825
L s= (1− ρ)
= 1−0.825
= 5trains(approximately).
Also,
P(≥ 10) = 0.82510 = 0.1460 .
Problem 2.
132 SRMIST DDE Self Instructional Material
Customers arrive at a window drive-in bank according to Poisson
NOTES
distribution with mean 10 per hour. Service time per customer is exponential
with mean 5 minutes. The space in front of the window including that for the
services car can accommodate a maximum of 3 cars. Other can wait outside
this space.
What is the probability that an arriving customer can drive directly to the
space in front of the window?
What is the probability that an arriving customer will have to wait outside the
indicated space?
How long the arriving customer is expected to wait before starting service?
Solution.
We know that
𝜆 𝑛 𝜆
𝑝𝑛 = µ
1−µ .
The probability that an arriving customer can drive directly to the space in
front of the window = 𝑝0 + 𝑝1 + 𝑝2
𝜆
𝑝0 = 1 − µ ;
𝜆 1 𝜆
𝑝1 = 1− ;
µ µ
𝜆 2 𝜆
𝑝2 = µ
1−µ ;
So,
𝜆 𝜆 1 𝜆 2
𝑝0 + 𝑝1 + 𝑝2 = 1 − µ 1+ µ
+ µ
We have,
𝜆 = 10per hour,
1
µ = 5 × 60 = 12per hour
𝜆 10
and ρ = µ = 12 .
Now,
133 SRMIST DDE Self Instructional Material
𝜆 𝜆 1 𝜆 2
𝑝0 + 𝑝1 + 𝑝2 = 1 − 1+ +
µ µ µ
NOTES
10 10 1 10 2
= 1 − 12 1+ 12
+ 12
= 0.42
Probability that an arriving customer will have to wait outside the indicated
space is equal to 1- 0.42 = 0.58 .
Average waiting time of a customer in the queue
𝜆 1 10 1
=
µ µ−𝜆 12 12 − 10
5
= = 0.417 hours .
12
Problem 3.
In a public Telephone booth the arrivals are on the average 15 per hour. A
call on the average takes 3 minutes. If there is just one phone find the
following expected number of callers in the booth at any timeandThe
proportion of the time the booth is expected to be idle?
Solution.
Mean arrival rate
𝜆 = 15per hour;
Mean service rate
1
µ= 3
× 60 = 20per hour.
Expected length of the non empty queue is
µ 20
= =4
µ − λ 20 − 15
The service is busy means
𝜆 15 3
µ
= 20
= 4.
3 1
Therefore the booth expected to idle for1− 4 = 4
hrs which is equal to
15minutes.
134 SRMIST DDE Self Instructional Material
Problem 4.
NOTES
A T.V repairman finds that the time spent on his job has an exponential
distribution with mean 30 minutes. If he repairs sets in the order in which
they came in and if the arrival of sets is Poisson with an average rate of 10 per
8 hour day, what is his expected idle time day? How many jobs are ahead of
the average set just brought in?
Solution.
Mean arrival rate
10
𝜆= 8
per hour;
Mean service rate
1
µ= 30
× 60 = 2 sets per hour.
The utilisation factor
10
𝜆 5
ρ= = 8
= .
µ 2 8
For 8 hour in a day, Repairman’s busy time is
5
8 X = 5 hours.
8
Idle time of repairman is 8 – 5 = 3 hours.
The number of jobs ahead = no of units in the system
ρ
L s= (1− ρ)
.
5
ρ 5
L s= = 8
5 = = 2 (approximately).
(1− ρ) (1− ) 3
8
Problem 5.
Cars arrive at a petrol pump, having one petrol unit, in Poisson fashion with
an average of 10 cars per hour. The service time is distributed exponentially
with a mean of 3 minutes. Find
Average number of cars in the system
Average waiting time in the queue
Average queue length
135 SRMIST DDE Self Instructional Material
The probability that the number of cars in the system is 2
NOTES
Solution.
Mean arrival rate
𝜆 = 10per hour;
Mean service rate
1
µ= 3
× 60 = 20 cars per hour.
The utilisation factor
𝜆 10 1
ρ= = = .
µ 20 2
Average cars in the system
1
ρ
L s= = 2
1 = 1.
(1− ρ) (1− )
2
Average waiting time in the queue
Lq 0.5
= = 0.05 .
𝜆 10
Average queue length queue
1
ρ2 1
L q= = 4
1 = = 0.5 car.
(1− ρ) 1− 2
2
Probability on n units in the system is
𝑛
𝜆 𝜆
𝑝𝑛 = 1−
µ µ
When n = 2,
1 2 1 1
𝑝2 = 2
1− 2
= 8
.
Problem 6.
The rate of arrival of customers at a public telephone booth follows Poisson
distribution with an average time of 10minutes between one customer and
the next.
136 SRMIST DDE Self Instructional Material
The duration of a phone call is assumed to follow exponential distribution,
NOTES
with mean time of 3 minutes.
What is the average probability that a person arriving at the booth will have
to wait?
What is the average length of the nonempty queues that form from time to
time?
Estimate the fraction of a day that the phone will be in use.
What is the probability that it will take him more than 10 minutes altogether
to wait for phone and complete his call?
Solution.
Mean arrival rate
1
𝜆= × 60 = 6per hour.
10
Mean service rate
1
µ= 3
× 60 = 20 per hour.
The probability that a person arriving at the booth will have to wait
𝜆 6
P(w > 0) = 1 – P0 = 1 − µ
= 20
= 0.3
Average length of non-empty queues
µ 20
= = 1.43
(µ − λ) 20 − 6
The fraction of a day that the phone will be busy
= the traffic intensity
𝜆 6
= ρ= = = 0.3
µ 20
∞
𝜆 − µ −λ 𝑤
P(w ≥ 10) = 𝜆 1− 𝑒 𝑑𝑤
µ
10
= (0.30)(0.23) 𝑒 −0.23𝑤 𝑑𝑤
10
Where 𝜆 = 0.10 per minute, and µ = 0.33 per minute.
137 SRMIST DDE Self Instructional Material
Therefore
NOTES 𝑒 −0.23
P w ≥ 10 = 0.069 −0.23
,
Applying the lower limit 10 and upper limit ∞ ,
we get P w ≥ 10 = 0.03.
This shows that 3 percent of the arrivals on an average will have to wait for
10 minutes or more before they can use the phone.
Problem 7.
On an average 96 patients per 24 hour day require the service of an
emergency clinic. Also on an average, a patient requires 10 minutes of active
attention.
Assume the facility can handle only one emergency at a time. Suppose that it
costs the clinic Rs. 100 per patient treated to obtain an average servicing time
of 10 minutes, and that each minute of decrease in this average time would
cost Rs 10 per patient treated.
How much would have to be budgeted by the clinic to decrease the average
1 1
size of the queue from 1 3
patients to 2
a patient.
Solution.
The average number of patients in the queue are given by
2 2
ρ2 4
L q =(1− ρ) = 3
2 = .
1− 3
3
Fraction of the time for which there are no patients given by,
𝜆 2 1
P0 = 1 − ρ = 1 − µ
=1− 3
= 3
.
4 1
Now, when the average queue size is decreased from 3
patients to 2 patient,
we are to determine the value of µ. So, we have
Λ2 1
=
µ µ− λ 2
This implies
138 SRMIST DDE Self Instructional Material
(1/15)2 1
= impliesµ = 2/15 NOTES
µ(µ−( 1/15)) 2
Therefore the average rate of treatment required
1 15
= = 7.5minutes.
µ 2
A decrease in average rate of treatment is (10 - 7.5) minutes or 2.5 minutes.
Budget per patient
Rs. (100 + 2.5 X 10 ) = Rs. 125.
Hence, inorder to get the required soze pf the queue, the budget should be
increased from Rs. 100 per patient to Rs. 125 per patient.
Exercise Problems
Problem 1. A supermarket has a single cashier. During the peak hours,
customers arrive at a rate 20 customers per hour. The average number of
customers that can be processed by the cashier is 24 per hour. Calculate:
1. Probability that the cashier is idle.
2. Average number of customers in the queuing system.
3. Queue size.
4. Average time a customer spends in the system.
5. Average time a customer spends in the queue waiting for service.
Solution: 0.167; 5 customers ; 4 customers ; 15minutes ; 12.5 minutes
Problem2.
Customers arrive at a sales counter manned by a single person according to
Poisson process with a mean rate of 20 per hour.
The time required to serve a customer has an exponential distribution with a
mean of 100 seconds. Find the average waiting time of a customer.
Solution: waiting time in queue = 125 sec ;wating time in system = 225 sec
Problem 3.
139 SRMIST DDE Self Instructional Material
A repair shop has a single mechanic to attend customers who bring small
NOTES appliances for repairs. The mechanic inspects the appliances for defects and
fixes them right away or renders a diagnosis.
All this takes him 6 minutes on an average. If 4 customers visit the shop per
hour on an average and if arrivals and service times are exponential,
determine:
1. Proportion of time during which the shop is empty;
2. Probability of finding at least one customer in the shop;
3. Average number of customers in the shop;
4. Average time spent by a customer in the shop.
Solution: 0.60 ; 0.40 ; 2/3 customers approx. ; 1/6 hours.
140 SRMIST DDE Self Instructional Material
______________________________________________________________ NOTES
4.2 MODEL II - (M/M/1) : (𝑵/FIFO)
The probability of n arrivals given by
(1− ρ)ρ𝑛
Pn= where ρ ≠ 1 ; 0 ≤ n ≤ N .
(1−ρ𝑁+1 )
1
Pn= where (ρ = 1)
𝑁+1
The relationship between Pn and P0 is given by
Pn = P0ρ𝑛
Average number of customers in the system is given by
ρ 1 − N + 1 ρ𝑁 + Nρ𝑁+1
𝐸𝑛 =
1 − ρ (1 − ρ𝑁+1 )
Average queue length is given by
ρ2 1 − Nρ𝑁−1 + (N − 1)ρ𝑁
𝐸𝑚 =
1 − ρ (1 − ρ𝑁+1 )
The average waiting time in the system is given by
E(n)
E(v) =
λ(1−𝑃𝑁 )
The average waiting time of the queue can be obtained by
E(m) 1
E(w) = (or) E(w) = E v − µ
λ(1−𝑃𝑁 )
Problem1.
At a railway station only one train is handled at a time. The railway yard is
sufficient only for two trains to wait while other is given signal to leave the
station. Trains arrive at the station at an average rate of 6 per hour and the
railway station can handle them on an average or 12 per hour.
Assuming Poisson arrivals and exponential service distribution, find the
steady-state probabilities for the various number of trains in the system.
Also find the average waiting time of a new train coming into the yard
141 SRMIST DDE Self Instructional Material
Solution.
NOTES
Given λ = 6 and µ = 12
so that
6
ρ = 12 = 0.5 .
The maximum queue length is 2, i.e.,
The maximum number of trains in the system is 3.
The probability that there is no train in the system(both waiting and in
service) is given by
1− ρ ρ𝑛
Pn= 1−ρ𝑁+1
1− 0.5 0.5 0
= (1−0.53+1 )
= 0.53 (since N=3, n=0).
We know that
Pn = P0ρ𝑛 therefore we have
P1 = (0.53) (0.5) = 0.27,
P2 = (0.53) (0.5)2 = 0.13
and P3 = (0.53) (0.5)3 = 0.07
Hence, we get
E(n) = 1(0.27) + 2(0.13) + 3(0.07) = 0.74.
Thus the average number of trains in the system is 0.74 and each train takes
on an average 0.08 hours for getting service.
As the arrival of new train expects to find an average of 0.74 trains in the
system before it.
E(w) = (0.74)(0.08) hours = 0.0592 hours.
Problem2.
Assume that the goods trains are coming in a yard at the rate of 30 trains per
day and suppose that the inter arrival times follow an exponential
distribution.
142 SRMIST DDE Self Instructional Material
The service time for each train is assumed to be exponential with an average
NOTES
of 36 minutes. If the yard can admit 9 trains at a time (there being 10 lines,
one of which is reserved for shunting purposes), calculate the probability that
the yard is empty and find the average queue length.
Solution.
We have Mean arrival rate
30 1
𝜆= 60×24
= 48
per minute
Mean service rate
1
µ= per minute
16
𝜆 36
and ρ = µ = 48
= 0.75.
The probability that the yard is empty is given by
1− ρ
P0= 1−ρ𝑁+1
1−0.75 0.25
= 1−0.75 10
= 0.90
= 0.28.
Average queue length is given by
ρ2 1 − Nρ𝑁−1 + (N − 1)ρ𝑁
𝐸𝑚 =
1 − ρ (1 − ρ𝑁+1 )
0.752 1 − 9 0.75 8 + 8 0.75 9
=
1 − 0.75 (1 − 0.7510 )
1 − 0.303
= 2.22
(1 − 0.005 )
= 2.22 0.70 = 1.55
Exercise
Problem1.
Consider a single queuing system with Poisson input, exponential service
times. Suppose the mean arrival rate is 3 calling units per hour, the expected
service time is 0.25 hours and the maximum permissible number of calling
units in the system is two. Derive the steady-state probability distribution of
143 SRMIST DDE Self Instructional Material
the number of calling units in the system, and then calculate the expected
NOTES number in the system.
Solution: Pn = 0.4 = 0.43(0.75)n ; P0 = 0.432 and E(n) = 0.81
Problem2.
If for period of 2 hours in the day trains arrive at the yard every 20 minutes
but the service time continues to remain 36 minutes, then calculate for this
period
The probability that the yard is empty
The average number of trains in the system; on the assumption that the line
capacity of the yard is limited to 4 trains only.
Solution: P0 = 0.04 ; E(n) = 3 trains
Problem3.
Patients arrive at a clinic according to a Poisson distribution at a rate of 30
patients per hour. The waiting toon does not accommodate more than 14
patients. Examination time per patient is exponential with mean rate 20 per
hour.
Find the effective arrival rate at the clinic.
What is the probability that an arriving patient will not wait?
What is the expected waiting time until a patient is discharged form the
clinic?
Solution: 20 patients(approx.) ; 0.67 ; 0.65
Problem4.
In a car wash service facility, cars arrive for service according to aPoisson
distribution with mean 5 per hour. The time for washing and cleaning each
car varies but is found to follow an exponential distribution with mean 10
minutes per car. The facility cannot handle more than one car at a time and
has a total of 5 parking spaces.
144 SRMIST DDE Self Instructional Material
Find the effective arrival rate
NOTES
What is the probability that an arriving car will get service immediately upon
arrival?
Find the expected number of parking spaces occupied.
Solution: 4.5 cars ; 0.17 ; 1.2
145 SRMIST DDE Self Instructional Material
CHAPTER V GAME THEORY
NOTES
STRUCTURE
5.0 GAME THEORY CONCEPTS
Game
A competitive situation will be called a Game, if it has the following
properties.
There are finite number of competitors called players.
Each player has a finite number of strategies available for him.
A play of the takes place when each player employs his strategy.
Every game results in an outcome loss or gain or a draw usually
called payoff to some player.
Strategy.
A strategy for a player is defined as a set of rules or alternative courses of
action available to him in advance, by which player decides the course of
action that he should adopt. A strategy may be of two types.
Pure Strategy
If the player select the same strategy each time then it is referred to as pure
strategy. In this case each player know exactly what the other player is going
to do, the objective of the players is to maximize or to minimize losses.
Mixed strategy
When the players use a combination of strategies and each player always
kept guessing as to which course of action is to be selected by the other player
at a particular occasion then this is known as mixed strategy.
146 SRMIST DDE Self Instructional Material
Optimum strategy NOTES
A course of action or play which puts the player in the most preferred
position, irrespective of the strategy of his competitors, is called an optimum
strategy.
Pay-off Matirx
The gains resulting from a two-person zero-sum game can be represented in
the matrix form usually called pay-off matrix.
Two person – zero sum game
A game with only two players is known as two – person zero sum game if the
one players gain is equal to the loss of other player such a way that total sum
is zero.
Saddle point
A saddle point of a pay-off matrix is that position in the ay-off matrix where
maximum of row minima coincides with the minimum of the column
maxima.
Value of a game
The value of a game is the maximum gain to the maximising player if both
the players use best strategy.
Problem1.
Find the value of game and saddle point of the following game.
147 SRMIST DDE Self Instructional Material
NOTES
Min(column max) = 50.
Max(row min) = 50
Min(column max) = Max(row min).
So, the value of the game is v= 50. Saddle point (1,3).
Problem 2. Solve the game whose pay-off matrix is given by
Solution:
Min(column max) = 1.
Max(row min) = 1.
148 SRMIST DDE Self Instructional Material
Min(column max) = Max(row min).
NOTES
So, the value of the game is v= 1.
Saddle point (1,1), (1,3) and (3,1).
149 SRMIST DDE Self Instructional Material
Problem 3.
NOTES
Determine the range of value of P and q that will make the payoff element a22
a saddle point for the game whose pay-off matrix is given below
Solution:
Ignoring p and q, we get
Value of game is 7.
Saddle point (2,2).
Row minimum of the second row is 7 after ignoring q value so q ≥7.
Column maximum of the second column is 7 after ignoring p value so p≤7.
Problem 4.
For what value of λ the game is strictly determinable?
150 SRMIST DDE Self Instructional Material
Solution:
NOTES
Min (column max) = -1. Max (row min) =2
So, when -1 <= λ <= 2 the game is strictly determinable.
151 SRMIST DDE Self Instructional Material
5.1ODD’S METHOD
NOTES
For solving 2 x 2 game, without saddle point, the following formula is used
If the pay-off matrix
a11 a22 – a21 a12
𝑉𝑎𝑙𝑢𝑒 𝑜𝑓 𝑡𝑒 𝑔𝑎𝑚𝑒 𝑣 =
(a11 + a22) – (a12 + a21)
a22 − a21
p1= (a11+ a22)−( a12+ a21) ;
a22 − a21
q1= (a11+ a22)−( a12+ a21) ;
p2 = 1 - p1 ;
q2 = 1 – q1.
Problem 1. Solve the following game.
Solution.
Min (column max) = 4.
Max (row min) =2.
Min (column max) ≠ Max (row min) .
152 SRMIST DDE Self Instructional Material
Saddle point does not exist.
NOTES
So, use odds method to solve the problem.
a11 a22 – a21 a12
𝑉𝑎𝑙𝑢𝑒 𝑜𝑓 𝑡𝑒 𝑔𝑎𝑚𝑒 𝑣 =
a11 + a22 – a12 + a21
1x2−5x4
= =3
(1+2) – (5+4)
a22 − a21
p1= a11+ a22−( a12+ a21)
2− 4 1
= = ;
−6 3
p2= 1 - p1
1 2
= 1-3=3;
a22 − a12
q1= a11+ a22−( a12+ a21)
2−5 1
= −6
= 2;
1 1
q2= 1 – q1 = 1 - =
2 2
Solving games by Dominance Method
If the game is not having saddle point we can use dominance method to solve
a game.
Dominance rules:
If all the elements of ith row less than or equal to the corresponding each
element of jth row, we can say that the ith row is dominated by the jth row.
Then eliminate ith row.
If all the elements of ith column less than or equal to the corresponding each
element of jth column, we can say that the jth column is dominated by the ith
column. Then eliminate jth column.
Problem 1.
153 SRMIST DDE Self Instructional Material
Solve the following game.
NOTES
Solution.
In this problem 2nd row (10, 6, 2) is smaller than 3rd row (20, 15, 18).
So the 2nd row is dominated by the 3rd row.
So, eliminate 2nd row. The reduced matrix is given by
The first column is greater than the third column so eliminate the column 1.
The reduced matrix is given by
No further comparison is possible.
We reduced the given 3X3 matrix into 2X2 matrix.
154 SRMIST DDE Self Instructional Material
We further use odds method to solve the game.
NOTES
a11 a22 – a21 a12
𝑉𝑎𝑙𝑢𝑒 𝑜𝑓 𝑡𝑒 𝑔𝑎𝑚𝑒 𝑣 =
(a11 + a22) – (a12 + a21)
(20x18)+(10x15) 170
= (20+18) – (10+15)
= 11
.
Probability of A selecting strategies:
The probability of selecting the strategy A1 is
a22 − a21 3 1
p1= a11+ a22−( a12+ a21) = 33
= 11
;
The probability of selecting the strategy A2 is
p2 = 0 ( Since we eliminated A2)
The probability of selecting the strategy A3 is
1 2
p3= 1 - p1 = 1 - 3 = 3.
The probability of B selecting the strategies:
The probability of selecting the strategy B1 is
q1 = 0 (since we eliminated B1)
The probability of selecting the strategy B2 is
a22 − a12 18+10 28
q2= a11+ a22−( a12+ a21) = 33
= 33 ;
The probability of selecting the strategy B3
28 5
q3 = 1 – q1 = 1 - 33 = 33 ;
Graphical method for solving game:
To use graphical method the matrix should be 2 X n or m X 2 matrix.
We use this method to reduce 2 X n or m X 2 matrix into 2 X 2 matrix.
If the matrix is 2 X n matrix then we have to select the lower envelope in the
graph and the identify the maximum point in the lower envelope.
If the matrix is m X 2 matrix then we have to select the upper envelope in the
graph and the identify the minimum point in the upper envelope.
Finally, we use odds method to find value of the game.
155 SRMIST DDE Self Instructional Material
Problem 1.
NOTES
Use graphical method for solving the following game.
Solution.
Max(row minimum) ≠ Min(column maximum).
Saddle point does not exist.
Also, no column or row dominated by other column or row. So we cannot
use dominance method to reduce the matrix.
The given matrix is 2 X n matrix so it is suitable to use graphical method to
reduce the matrix.
156 SRMIST DDE Self Instructional Material
NOTES
The reduced matrix which corresponding to the strategies B3 and B4 is given
by
The value of the game
a11 a22 – a21 a12 22
𝑣= (a11+ a22) – (a12 + a21)
= 9
.
Probability of A selecting strategies:
The probability of selecting the strategy A1 is
4
p1= 9
157 SRMIST DDE Self Instructional Material
The probability of selecting the strategy A3 is
NOTES 4 5
p2= 1 - p1 = 1 - 9 = 9 .
The probability of B selecting the strategies:
The probability of selecting the strategy B1 is
q1 = 0 (since we eliminated B1).
The probability of selecting the strategy B2 is
q2 = 0 (since we eliminated B2)
The probability of selecting the strategy B3 is
a22 – a12 8
q3= a11+ a22− a12+ a21
=9 .
The probability of selecting the strategy B4 is
8 1
q4 = 1 – q4 = 1 - = .
9 9
Problem 2. Solve the following game graphically.
158 SRMIST DDE Self Instructional Material
Solution.
NOTES
The game is an mX2 game so we selected the upper envelope and the given
matrix reduced into the 2 x 2 matrix.
The value of the game
a11 a22 – a21 a12 73
𝑣= (a11+ a22) – (a12 + a21)
= 17 .
Probability of A selecting strategies:
The probability of selecting the strategy A1 is
p1 = 0 (since we eliminated A1).
159 SRMIST DDE Self Instructional Material
The probability of selecting the strategy A2 is
NOTES a22 – a21 4
p2 = a11+ a22− a12+ a21
= 17 .
The probability of selecting the strategy A3 is
4 13
p3= 1 - p2= 1 - 17 = 17 .
The probability of selecting the strategy B1 is
p4 = 0 (since we eliminated A4).
The probability of B selecting the strategies:
The probability of selecting the strategy B1 is
a22 – a12 5
q1= a11+ a22− a12+ a21
= 17 .
The probability of selecting the strategy B2 is
12
q2= 17 .
Note: The strategies A1,A4 and A5 are dominated by the strategy A2 . So
eliminate A1,A4 and A5 . So by using dominance property, we can reduce the
matrix into 2 X 2.
160 SRMIST DDE Self Instructional Material
5.2 MATRIX ODDMENT METHOD NOTES
It is suitable for n x n matrix for n=3.
If we cannot reduce the payoff matrix into 2X2 by using dominance method
that time we use oddment to reduce the matrix.
Let A be an nXn matrix. Obtain a new matrix C whose first column is A’s 1st
column – 2ndColumn and the 2nd column is A’s 2nd column – A’s 3rd column.
Obtain a new matrix R whose first row is A’s 1st row – 2ndrow and the 2nd
row is A’s 2nd row – A’s 3rd row.
Ci be the matrix by deleting ith row from C.
Rj be the matrix obtained by deleting jth column from R.
Find the determinants of Ci’s and Rj’s.
Problem 1.
Solve the following 3X3 game
Solution.
No row or column dominated by other rows or column.
So we cannot reduce the matrix in to 2x2 or 2xn or mxn order by dominance
method.
161 SRMIST DDE Self Instructional Material
NOTES
162 SRMIST DDE Self Instructional Material
NOTES
│C1│ = 40 ; │C2│ = -22 ; │C3│ = 28
│R1│ = 28 ; │R2│ = -22 ; │R3│ = 40
The value of the game is
40 22 28 −29
𝑣 = 90 X 3 + 90 X -3 + 90 X -4 = 45
The probability of A selecting the strategies
The probability of selecting the strategy A1 is
40
p1= 90 .
The probability of selecting the strategy A2 is
22
p2 = 90 .
The probability of selecting the strategy A3 is
28
p3= 90
.
The probability of B selecting the strategies:
The probability of selecting the strategy B1 is
28
q1= .
90
The probability of selecting the strategy B2 is
163 SRMIST DDE Self Instructional Material
22
q2 = .
90
NOTES
The probability of selecting the strategy B2 is
40
q3= .
90
Exercise problems.
1. Two companies A and B competing for the same product. Their
different strategies are given in the following payoff matrix. Solve the
game.
Answer.
A1 = 9/14; A2 = 5 /14 ; B1 = 2/3 ;B2 = 1/3 ;V =1/3.
2. Solve the following game
Answer. (3/8 ,13/24 , 1/12) ; (0 , 3/7 , 4/7 ) and V = 1/3.
3. Solve the game graphically.
Answer. (2/5 , 3/5) (0 , 4/5 , 0 , 1/5) and V = 2/5.
164 SRMIST DDE Self Instructional Material
4. Solve the Game graphically
NOTES
Answer. ( 0 , 3/5 , 0 , 2/5 , 0 , 0) (4/5 , 1/5) and V = 17/5.
5. Use dominance property and graphical method to solve the problem.
Answer. ( 0 , 0 , 2/3 , 1/3) (0 , 0, 2/3 , 1/3) and V = 8/3.
165 SRMIST DDE Self Instructional Material
5.3 SIMULATION
NOTES
Definition
Simulation is the process of designing a model of a real system and
conducting experiments with this model for the purpose of understanding
the behaviour for the operation of the system
Why we need 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.
Would be disruptive.
Would be too expensive.
Does not permit replication of events.
Does not permit control over key variables.
2. It is a desirable tool for solving a business problem when a
mathematical model is
Too complex to solve,
Beyond the capacity of available personnel
Not detailed enough to provide information
3. The major reasons for adopting simulation in place of other
mathematical techniques are
Only one method available
Without appropriate assumption it is impossible to develop
solution
Too expensive to observe system
4. It provides a trial and error movement towards the optical solution.
The decision maker selects an alternative, experiences the effect of the
selection, and then improves the selection.
166 SRMIST DDE Self Instructional Material
5.4 MONTE – CARLO SIMULATION NOTES
The Monte-Carlo method is a simulation technique in which statistical
distribution functions are created by using a series of random numbers.
Procedure
Step 1.
Define the problem
Identify the objectives of the problem, and identify the main factors which
have the greatest effect on the objectives of the problem
Step 2
Construct an appropriate method
Specify the variables and parameters or the model
Formulate the appropriate decision rules, state the conditions under which
the experiment is to be performed.
Identify the type of distribution that will be used – Models use either
theoretical distributions or empirical distributions to state the patterns the
occurrence associated with the variables.
Specify the manner in which time will change.
Define the relationship between variables and parameters.
Step 3.
Prepare the model for experimentation
Define starting conditions for the simulation and specify the number of runs
of simulation to be made
Step 4.
UsingStep 1 and Step 3 experiment with the model
Define a coding system that will correlate the factors defined in Step 1 with
the random numbers to be generated for the simulation
Select a random number generator and create the random numbers to be used
in the simulation
167 SRMIST DDE Self Instructional Material
Associate the generated random numbers with the factors identified in Step
NOTES 1and coded in Step 4.
Step 5.
Summarize and examine the results obtained in Step 4.
Step 6.
Evaluate the results of the simulation
Step 7.
Formulate proposals for advice to management on the course of action to be
adopted and modify the model, if necessary.
Problem1.
The automobile company manufactures around 150 scooters. The daily
production varies from 146 to 154 depending upon the availability of raw
materials and other working conditions
The finished scooters are transported in a specially arranged lorry
accommodating 150 scooters.
Using the following random numbers.
80, 81, 76, 64, 43, 18, 26, 10, 12, 65, 68, 69, 61, 57.
Simulate the process to find out
What will be the average number of scooters waiting in the factory?
What will be the average number of empty space on the lorry?
Solution.
The random numbers are established as in table below
168 SRMIST DDE Self Instructional Material
NOTES
Based on the random numbers given, we simulate the production per day in
the table below.
169 SRMIST DDE Self Instructional Material
NOTES
The average number of scooters waiting 21/5 = 1.4 per day
Average number of empty spaces 9/ 15 = 0.6 per day.
Learning Resources:
Text books
1) Hamdey A. Taha, Introduction to operations research, Prentice Hall
India, Tenth edition, 2017
2) Fredericks, Hiller, Gerald J. Lieberman Bodhibrata Nag Prectambasu,
Operations research 9e, Mcgraw hill education, 9th edition, 2017.
References
1. Kantiswarup, P. K. Gupta, Man mohan, Introduction to
Management science Operations research, Sultan chand and sons ,
2017.
2. Schaum's Outline of Operations Research (by Richard Bronson)
3. Introduction to Operations Research (by F S Hiller and G J
Leiberman)
4. Operations Research: Applications and Algorithms (by Wayne L.
Winston)
5. An Introduction to Management Science: Quantitative
Approaches to Decision Making (by David R. Anderson)
170 SRMIST DDE Self Instructional Material
You might also like
- Unit - V Linear Probramming - Assignment SheetNo ratings yetUnit - V Linear Probramming - Assignment Sheet4 pages
- Lecturer: R.Rajaraman Subject: Applied Operation Research Code: BA1603 UNIT I: Introduction To Linear Programming Part-A100% (1)Lecturer: R.Rajaraman Subject: Applied Operation Research Code: BA1603 UNIT I: Introduction To Linear Programming Part-A19 pages
- Mathematical Formulation of Linear Programming ProblemsNo ratings yetMathematical Formulation of Linear Programming Problems5 pages
- Operations Research Lecture 1: Introduction To OR Models: Kusum Deep Mathematics DepartmentNo ratings yetOperations Research Lecture 1: Introduction To OR Models: Kusum Deep Mathematics Department44 pages
- EContent 3 2024 09 18 17 33 13 Session416LinearProgrammingIntroduction2ppt 2024 08 16 14 07 54No ratings yetEContent 3 2024 09 18 17 33 13 Session416LinearProgrammingIntroduction2ppt 2024 08 16 14 07 5440 pages
- Linear Programming: Tuesday, October 22, 2019 1 Dr.G.Suresh Kumar@KL UniversityNo ratings yetLinear Programming: Tuesday, October 22, 2019 1 Dr.G.Suresh Kumar@KL University57 pages
- 9872 - Intro To LPP and Its FormulationNo ratings yet9872 - Intro To LPP and Its Formulation26 pages
- 2 Linear Programming (LP) Problem - FormulationsNo ratings yet2 Linear Programming (LP) Problem - Formulations57 pages
- Linear Programme-Simplex and Graphical PDF100% (1)Linear Programme-Simplex and Graphical PDF34 pages
- Student Solutions Manual to Accompany Loss Models: From Data to Decisions, Fourth EditionFrom EverandStudent Solutions Manual to Accompany Loss Models: From Data to Decisions, Fourth Edition4/5 (1)
- AI for Good: Applications in Sustainability, Humanitarian Action, and HealthFrom EverandAI for Good: Applications in Sustainability, Humanitarian Action, and HealthNo ratings yet
- 지원금 및 지원 수준 (Details on the Research Assistantship and Support) 교수 소속 및 연구분야 (Professor's Contact Details and Fields of Study)No ratings yet지원금 및 지원 수준 (Details on the Research Assistantship and Support) 교수 소속 및 연구분야 (Professor's Contact Details and Fields of Study)1 page
- Computer Aided Engineering Drawing: Entity Draw Commands in AutocadNo ratings yetComputer Aided Engineering Drawing: Entity Draw Commands in Autocad37 pages
- Cyber Security Workshop Lab File CompleteNo ratings yetCyber Security Workshop Lab File Complete43 pages
- Furniture Showroom Management System Project Report.No ratings yetFurniture Showroom Management System Project Report.44 pages
- 2022 - A Review - Data Pre-Processing and Data Augmentation Techniques - ScienceDirectNo ratings yet2022 - A Review - Data Pre-Processing and Data Augmentation Techniques - ScienceDirect20 pages
- Server Security Audit Report FOR: Baramati AgroNo ratings yetServer Security Audit Report FOR: Baramati Agro9 pages
- Impact of Technology On Business CommunicationNo ratings yetImpact of Technology On Business Communication3 pages
- Electromagnetic Clutch System Full Seminar Report 4537No ratings yetElectromagnetic Clutch System Full Seminar Report 453730 pages
- A Review of Methods, Techniques and Tools For Project Planning and ControlNo ratings yetA Review of Methods, Techniques and Tools For Project Planning and Control20 pages
- Lecture-2,3 - Chapter 2 - Organizing and Graphing DataNo ratings yetLecture-2,3 - Chapter 2 - Organizing and Graphing Data46 pages
- Toward Understanding Outcomes Associated With Data Quality ImprovementNo ratings yetToward Understanding Outcomes Associated With Data Quality Improvement12 pages
- Resource Estimation: Professor Emmanuel Chanda Curtin University of Technology/UNZA, SCH of MinesNo ratings yetResource Estimation: Professor Emmanuel Chanda Curtin University of Technology/UNZA, SCH of Mines32 pages
- De La Salle Santiago Zobel School High School Department Unit Assessment MatrixNo ratings yetDe La Salle Santiago Zobel School High School Department Unit Assessment Matrix6 pages