Or All Units Lecture Notes
Or All Units Lecture Notes
OPERATIONS RESEARCH
Introduction
It is the systematic method oriented study of the basic structure, characteristics, functions
and relationships of an organization to provide quantitative basis for decision making. OR is
concerned with scientifically deciding how to best design and operate man machine systems
usually requiring the allocation of Resources.
Some of the techniques of today's OR have been actually developed and used even
before the term OR was coined, some of the techniques are Inventory, Queuing Theory,
Statistical quality control etc.
After the World war-II the industries in America & Britain concentrated in applying
the Operations research methodologies to industrial problems for maximizing the profit with
limited Resources.
1) In Industry
OR techniques are useful in the field of production, inventory control, demand& forecast,
sales & purchase, transportation, repair & maintenance, scheduling & sequencing, planning,
scheduling & control of the projects.
2) In Defense
OR has a wide scope of application in defense operation. The activities performed by air
force, army, navy can be further divided into sub activities. i. e operations, intelligence,
administrates, and training etc. There is a need to co-ordinate various activities involved in order
to arrive optimum strategies and to achieve consistent Goal.
3) Planning
In both developing & developed economic OR approach is equally applicable. The basic
problem in more of the countries from Asia & Africa is remove poverty. Therefore there is a
great scope of Economics, Statistics, Administration, Technician &Agriculture experts working
together to solve the problem with an OR approach.
4) Agriculture
With population expansion & shortage of food every country is facing the problem of
optimum allocation of land with various crops in according to climate conditions and available
facilities.
5) Public utilities
OR Techniques can be applied in the area of transformation to regulate train arrivals and
their running times. OR methods can also applied to big hospitals to reduce waiting time of
patients. It has extensive uses in petroleum, chemical, paper, metal processing, air craft, rubber
mining and textile industries.
Phases of OR Study
The scientific method of OR study involves the following phases.
2) Research phase: This phase is the largest & longest among all the phases. This phase utilizes
3) Action phase:
This phase consists of making recommendations for implementing the decision. This
decision is implemented by an individual who is in a position to implement result. This
individual must aware of the environment in which the problem is occurred.
Objectives of OR
The objective of OR is to minimize the total cost and maximize the profit. Every
organization has a number of functional units or departments each perform a part of whole
job. With Economic growth, uncertainty is also growing; this makes each decision
complicate and time consuming.
In the competitive world today one has to take quick decision, because any delay may
help the competitors. The decision has to be quick as well as sound & requires a scientific
approach to the problem. The application of OR methods to helps in making the decisions in
such complex situations. OR combines the knowledge of various disciplines such as
mathematics, statistics, economics & engineering.
Objective of OR is to provide a scientific basis to the managers of an organization for
solving problems involving components of the system by employing a team of scientists
drawn from different disciplines.
Models of OR
There is no set of unique problems which can be solve by using OR models or techniques.
Several OR models or techniques can be grouped into some basic categories as given below.
1) Allocation model
2) Inventory model
3) Competitive model
4) Network model
6) Sequencing model
7) Replacement model
Uses of OR
• In all industries & other situations, resources are limited. OR studies help in effectively
selecting and distributing these resources.
• By using OR techniques the quality of decisions are better and in most of the cases
optimal.
• As it is a scientific way of making decisions, it gives a lot of scopes for today's manager
to develop skills or knowledge to arrive at optimal decisions.
• They give a number of decisions yielding the same results; hence depending upon the
situation most appropriate decision can be selected or used.
Limitations of OR
• Mathematical models which essence of OR, do not take into account qualitative factors or
emotions factor which are quite real. All influencing factors which cannot be quantified
find no place in mathematical models.
• OR tries to find optimal solution taking all factors of the problem into account. Present
day problems involve numerous such factors, expressing them in quantity and
establishing relations among them requires huge calculations.
• Being a new field generally there is a resistance from the employees to the new
proposals.
• Management who has to implement the proposals may itself offer a lot of resistance due
to conventional thinking.
• Young enthusiastic persons over taken by its advantages and exactness generally forget
that OR mean for men and not that men are meant for it.
Formulation of LPP
Given a situation describing the availability of resources and conditions and the object to be
optimized. we have to convert in to a mathematical model , this process is called formulation.
Step-1: Defining the Decision variables and writing the objective function.
Step-2: Expressing the conditions of availability of the resources in the form of linear
inequalities or equalities using the decision variables.
Subject to constraints:
𝒂𝟏𝟏 𝒙𝟏 + 𝒂𝟏𝟐 𝒙𝟐 + − − − − − − + 𝒂𝟏𝒏 𝒙𝒏 (≤, =, ≥) 𝒃𝟏
−−−−−−−−−−−−−−−−−−−−−−−−−
Processing Time
Product Profit
P1 P2
A(x1) 2 hrs. 1 hrs 20
B(x2) 4 hrs 3 hrs 12
Availability 20 hrs 23 hrs
𝒙𝟏 + 𝟑𝒙𝟐 ≤ 𝟐𝟑
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎 [𝒏𝒐𝒏 − 𝒏𝒆𝒈𝒂𝒕𝒊𝒗𝒆 𝒓𝒆𝒔𝒕𝒓𝒊𝒄𝒕𝒊𝒐𝒏𝒔]
2. National paints produce both interior and exterior paints from 2 raw materials M1 and
M2. The following table provides basic data of the problem.
A market survey indicates that the daily demand for interior paint cannot exceed that of
exterior paint by more than one ton, also maximum daily demand for interior paint is 2
tons. Formulate as LPP.
Solution. Let x1, x2 be the exterior and interior paints produce daily in tons. Assume Z be the
daily profit.
Subject to
𝒙𝟏 + 𝟐𝒙𝟐 ≤ 𝟔,
𝒙𝟐 − 𝒙𝟏 ≤ 𝟏,
𝒙𝟐 ≤ 𝟐
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
Exercise:
1. A firm manufactures 2 types of products A &B and sells them at a profit of Rs. 2 on type
A and Rs.3 on type B. Each product is processed on 2 machines G and H. Type A requires
1 minute processing time on G and 2 minutes on H. Type B requires 1 minute on G an 1
minute on H. The machine G is available for not more than 6hours 40 minutes, while H
available for 10 hours on any working day. Formulate this problem as LPP.
Solution of LPP
To solve LPP we use the following methods.
1) Graphical Method: If LPP consists of only 2 decision variables and whatever be the form of
constraints (≤, =, ≥) and inequalities we use graphical method.
2) Simplex Method: If LPP having 2 or more decision variables and if all the inequalities are in
the form of ≤, we use simplex method.
3) Artificial variable technique: If LPP having 2 or more variables and at least one constraint is
in the form of ≥, = we use artificial variable technique.
b) Big-M method
1) Graphical Method:
Procedure
ii) Plot each equation on the graph geometrically each one represents a straight line.
iii) Every point on the line will satisfies the equation of the line.
iv) If the inequality constraints corresponding to the line is ‘≤’. The region below the line in the
1st quadrant is shaded, for any inequality constraint with ‘≥’ the region above the line in 1st
quadrant is shaded, the points lying in the common region will satisfy all the constraints
simultaneously. The common region is called as “Feasible Region”.
vi) Locate the corner points of optimal solution by calculating the value of Z for each corner
point.
Subject to
𝒙𝟏 + 𝒙𝟐 ≤ 𝟒𝟎𝟎,
𝟐𝒙𝟏 + 𝒙𝟐 ≤ 𝟔𝟎𝟎,
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
Solution.
𝑥1 = 400
At (400, 0) equation (1) meets 𝑥1 -axis
𝑥2 = 400
At (0, 400) equation (1) meets 𝑥2 -axis
𝑥1 = 300
At (300, 0) equation (2) meets 𝑥1 -axis
𝑥2 = 600
At (0, 600) equation (2) meets 𝑥2 -axis
Here OABC is the feasible region where O(0,0), A(300,0), B(200,200),C(0,400)
Max Z at O(0,0) = 0
Max Z at A(300,0)=600
Max Z at B(200,200)=1000
Max Z at C (0,400)=1200
∴ Max Z =1200
And 𝑥1 = 0, 𝑥2 = 400
Subject to
𝟑𝒙𝟏 + 𝒙𝟐 ≥ 𝟑𝟎,
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
Solution.
3𝑥1 + 𝑥2 = 30 -------------(2)
𝑥1 = 40
At (40, 0) equation (1) meets 𝑥1 -axis
𝑥2 = 20
At (0, 20) equation (1) meets 𝑥2 -axis
𝑥1 = 10
At (10, 0) equation (2) meets 𝑥1 -axis
𝑥2 = 30
At (0, 30) equation (2) meets 𝑥2 -axis
𝑥2 = 20
At (0, 20) equation (3) meets 𝑥2 -axis
Here ABCD is the feasible region where A(15,0), B(40,0),C(4,18) ,D(6,12)
Here we get point C by solving (1) and (2) and by solving (2) and (3) we get D.
Min Z at A(15,0)=300
Min Z at B(40,0)=800
Min Z at D(6,12)=240
∴ Min Z =240
And 𝑥1 = 6, 𝑥2 = 12
General LPP
𝒁 = 𝒄𝟏 𝒙𝟏 + 𝒄𝟐 𝒙𝟐 + − − − − − − + 𝒄𝒏 𝒙𝒏
−−−−−−−−−−−−−−−−−−−−−−−−−
𝒂𝒏𝒅 𝒙𝒊 ≥ 𝟎 , i = 1,2,3, − − − − −n
Solution to LPP
A set of {𝑥1 , 𝑥2 , 𝑥3 , − − − − −𝑥𝑛 } satisfying the constraints of general LPP is called solution
to LPP.
Feasible Solution
Any solution which is also satisfies non-negative restriction of the problem is called feasible
solution.
Basic Solution
If there are m equality constraints and m+n is the number of variables(m≤n), a start for the
optimal solution is made by putting n unknowns(variables) equal to zero and then solving for m
equations in remaining m unknowns. The ‘n’ zero variables are called non basic variables and
remaining m variables are called basic variables which form a basic solution.
A Feasible solution which also a basic solution to LPP is called Basic Feasible Solution .It is
of two types.
i) Degenerate BFS: If any of the basic variables are zero then that solution is degenerate BFS.
ii) Non-Degenerate BFS: If all m variables are positive then that solution is Non-Degenerate
BFS.
Simplex Method
In solving LPP we find the set of non-negative values which satisfies the constraints and
objective function. If we have a set of m equations with (m+n) unknowns then an infinite set of
solutions are possible. A hit & trail for optimal solution is not feasible, there is one efficient &
systematic procedure which gives solution in a finite number of trails known as simplex method
or simplex technique.
Slack variable:
Surplus variable:
Let constraints of LPP are in the form
𝑀𝑎𝑥 𝑍 = 𝑐1 𝑥1 + 𝑐2 𝑥2 + − − − − − − + 𝑐𝑛 𝑥𝑛
and 𝑥1 , 𝑥2 , − − − − −𝑥𝑛 ≥ 0
• The objective function should be maximization type ,if it is minimization convert by using the
formula
Simplex Algorithm
The existence of initial basic feasible solution is always assumed the steps for competition of
an optimum solution given below.
Step-1: Check whether the objective function of the given LPP is to be maximized or minimized.
If it is to be minimized, we have to convert into maximization using the formula
Step-2: Check whether all 𝑏𝑖 (𝑖 =1,2,3,-------m) are positive ,if any 𝑏𝑖 is negative then multiply
the equation with -1 on both sides.
Step-3: Express the problem in standard form by introducing slack or surplus variables to convert
inequalities into equalities.
Step-4: Let 𝑋𝐵 be Initial Basic Feasible Solution to the LPP
𝑀𝑎𝑥 𝑍 = 𝐶𝑋
𝐴𝑋 = 𝑏
𝑋≥0
Such that it satisfies 𝑋𝐵 = 𝐵 −1 𝑏 , where b is the basis matrix formed by the basic variables.
𝑐1 𝑐2 𝑐3 𝑐𝑛 0
𝐶𝐵 𝐵𝑉 𝑋𝐵 𝑥1 𝑥2 𝑥3 𝑥𝑛 𝑥𝑛+1
i) If all (𝑍𝑗 − 𝐶𝑗 ) ≥ 0 then the initial basic feasible solution 𝑋𝐵 is an optimum basic feasible
solution.
If there are more than one negative (𝑍𝑗 − 𝐶𝑗 ). Choose the most negative of them.
Let (𝑍𝑘 − 𝐶𝑘 ) be the most negative of for j=k, this gives the entering variable 𝑋𝑘 and indicate it
by arrow ↑ at the bottom of kth column.
If there are more than one variable having the same most negative (𝑍𝑗 − 𝐶𝑗 ) . Choose any one of
them arbitrary as entering variable.
Step-8: For new basis by dropping the leaving variable and introducing the entering variable
along with associated value under 𝐶𝐵 column .convert the pivot element to 1 and the remaining
elements in that corresponding column to zero.
Step-9: Go to step-5 and repeat the process until either an optimum solution is obtained or there
is an indication of unbalanced solution.
Subject to
𝟐𝒙𝟏 + 𝟑𝒙𝟐 ≥ 𝟔,
𝟑𝒙𝟏 + 𝟒𝒙𝟐 ≤ 𝟖,
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
Answer.
𝟐𝒙𝟏 + 𝟑𝒙𝟐 − 𝒙𝟑 = 𝟔
If the inequality is ≤ type we in constraints we add Slack variable (Let it be 𝒙𝟒 ). Therefore the
second constraint becomes
𝟑𝒙𝟏 + 𝟒𝒙𝟐 + 𝒙𝟒 = 𝟖
𝟐𝒙𝟏 + 𝟑𝒙𝟐 − 𝒙𝟑 = 𝟔,
𝟑𝒙𝟏 + 𝟒𝒙𝟐 + 𝒙𝟒 = 𝟖
𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 , 𝒙𝟒 ≥ 𝟎
Subject to
𝟔𝒙𝟏 + 𝟒𝒙𝟐 − 𝒙𝟑 ≤ 𝟒,
𝒙𝟏 + 𝒙𝟐 + 𝒙𝟑 ≤ 𝟓,
𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 ≥ 𝟎
Answer.
𝑀𝑎𝑥 𝑧 / = −𝑀𝑖𝑛 𝑧
There fore
6𝑥1 + 4𝑥2 − 𝑥3 + 𝑥4 = 4,
𝑥1 + 𝑥2 + 𝑥3 + 𝑥6 = 5,
𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , 𝑥5 , 𝑥6 ≥ 0
Subject to
𝟐𝒙𝟏 + 𝒙𝟐 ≤ 𝟖,
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
2𝑥1 + 3𝑥2 + 𝑥3 = 12
2𝑥1 + 𝑥2 + 𝑥4 = 8
𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 ≥ 0
1 0 12 12
=[ ] [ ]=[ ]
0 1 8 8
𝑥3 = 12, 𝑥4 = 8
Simplex starting table is
𝑪𝒋 6 7 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 =
𝑿𝒌 𝑿𝟐
𝟏𝟐
0 𝒙𝟑 12 2 𝟑 1 0 𝟑
= 𝟒(Min)←
𝟖
0 𝒙𝟒 8 2 1 0 1 =𝟖
𝟏
Z=𝑪𝑩 𝑿𝑩 =0 0 0 0 0 𝒁𝒋
-6 -7 0 0 𝒁𝒋 − 𝑪𝒋
↑
Here 𝐶𝐵 is cost of basic variable
The intersection of leaving variable and entering variable is known as Pivot element (Indicated in
Red color).
𝑪𝒋 6 7 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 =
𝑿𝒌 𝑿𝟏
𝟒
7 𝒙𝟐 4 2/3 𝟏 1/3 0 𝟐/𝟑
=𝟔
𝟒
0 𝒙𝟒 4 4/3 0 -1/3 1 𝟒/𝟑
=𝟑←
𝑪𝒋 6 7 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 =
𝑿𝒌 𝑿𝟏
7 𝒙𝟐 2 0 1 1/2 -1/2
6 𝒙𝟏 3 1 0 -1/4 3/4
Z=𝑪𝑩 𝑿𝑩 =32 6 7 2 1 𝒁𝒋
0 0 2 1 𝒁𝒋 − 𝑪𝒋
∴ Max z=32, 𝒙𝟏 = 𝟑, 𝒙𝟐 = 𝟐
2. Solve the following LPP by Simplex method
Subject to
𝟑𝒙𝟏 − 𝒙𝟐 + 𝟐𝒙𝟑 ≤ 𝟕,
𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 ≥ 𝟎
3𝑥1 − 𝑥2 + 2𝑥3 + 𝑥4 = 7
−2x1 + 4x2 + x5 = 12
𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , x5 , x6 ≥ 0
1 0 0 7
12
=[0 1 0] [12] = [ ]
8
0 0 1 10
𝑥4 = 7, 𝑥5 = 12, 𝑥6 = 10
Simplex starting table is
𝑪𝒋 -1 3 -2 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 𝑥6 =
𝑿𝒌 𝑿𝟐
0 𝒙𝟒 7 3 -1 2 1 0 0 ------
1 0 𝟏𝟐
0 𝑥5 12 -2 4 0 0 =𝟑
𝟒
0 1 𝟏𝟎
0 𝑥6 10 -4 3 8 0 = 𝟑. 𝟑𝟑
𝟑
𝑧/ 0 0
𝒁𝒋
0 0 0 0
=𝑪𝑩 𝑿𝑩 =0
1 -3 2 0 0 0 𝒁𝒋 − 𝑪𝒋
↑
𝑪𝒋 -1 3 -2 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 𝑥6 =
𝑿𝒌 𝑿𝟏
0 𝒙𝟒 0 5/2 0 2 1 1/4 0 4
3 𝒙𝟐 3 -1/2 1 0 0 1/4 0 −−
0 𝑥6 1 -5/2 0 8 0 -3/4 1 −−
𝑧/ 3/4 0
𝒁𝒋
-3/2 3 0 0
=𝑪𝑩 𝑿𝑩 =9
-1/2 0 2 0 3/4 0 𝒁𝒋 − 𝑪𝒋
↑
Second iteration table is
𝑪𝒋 -1 3 -2 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 𝑥6 =
𝑿𝒌 𝑿𝟏
-1 𝒙𝟏 4 1 0 4/5 2/5 1/10 0
0 𝑥6 11 0 0 10 1 -1/2 1
𝑧/ 4/5 0
𝒁𝒋
-1 3 10 1
=𝑪𝑩 𝑿𝑩 =11
0 0 12/5 1/5 4/5 0 𝒁𝒋 − 𝑪𝒋
Here all 𝑍𝑗 − 𝐶𝑗 values are positive.
𝑧 / = 11
∴ 𝑍 = −11
∴ Min z=-11
𝒙𝟏 = 𝟒, 𝒙𝟐 = 𝟓, 𝒙𝟑 = 𝟎
Exercise
1. Solve the following LPP by Simplex method
Subject to
𝒙𝟏 + 𝒙𝟐 ≤ 𝟒,
𝒙𝟏 − 𝒙𝟐 ≤ 𝟐,
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
Subject to
𝒙𝟏 + 𝟒𝒙𝟐 ≤ 𝟏𝟎
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
LPP may have constraints of ≥ type .In such cases the starting simplex table will not contain
an identity matrix. We introduce a new type of variable called artificial variable. These variables
are imaginary and cannot have any physical meanings. These variables can be eliminated when
they left the basis.
Proceure:
Step-1: Express the LPP in standard form by introducing slack/surplus or artificial variables.
Assign zero co-efficient to surplus variables and large penalty to artificial variable in the
objective function.
Step-2: Solve modified LPP by simplex method at any iteration of any simplex method any one
of the following cases may arise.
I) If at least one artificial variable is present in the basis when all (𝑍𝑗 − 𝐶𝑗 ) ≥ 0 then the given
LPP does not have any solution.
II) If there is no artificial variable in the basis then the Simple method is continued until optimal
solution has been obtained.
Subject to
𝒙𝟏 + 𝒙𝟐 ≥ 𝟑
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
2𝑥1 + 3𝑥2 + 𝑥3 = 30
3x1 + 2x2 + 𝑥4 = 24
x1 + x2 − x5 + a1 = 3
𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , x5 , a1 ≥ 0
1 0 0 30 30
=[0 ] [
1 0 24 ] = [ 24]
0 0 1 3 3
𝑥3 = 30, 𝑥4 = 24, a1 = 3
𝑪𝒋 6 4 0 0 0 -M
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 a1 =
𝑿𝒌 𝑿𝟏
0 𝑥3 30 2 3 1 0 0 0 15
0 0 8
0 𝑥4 24 3 2 0 1
-1 1 𝟑←
-M a1 3 1 1 0 0
Z M -M
=𝑪𝑩 𝑿𝑩 = -M -M 0 0 𝒁𝒋
-3M
-M- M 0
-M-4 0 0 𝒁𝒋 − 𝑪𝒋
6
↑
𝑪𝒋 6 4 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 =
𝑿𝒌 𝑿𝟏
0 𝑥3 24 0 1 1 0 2 12
0 𝑥4 15 0 -1 0 1 3 5←
6 𝒙𝟏 3 1 1 0 0 -1 ---
Z -6
6 6 0 0 𝒁𝒋
=𝑪𝑩 𝑿𝑩 =18
0 2 0 0 -6 𝒁𝒋 − 𝑪𝒋
↑
Second iteration table is
𝑪𝒋 6 4 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 =
𝑿𝒌 𝑿𝟏
0 𝑥3 14 0 5/3 1 -2/3 0
0 𝑥5 5 0 -1/3 0 1/3 1
6 𝒙𝟏 8 1 2/3 0 1/3 0
Z=𝑪𝑩 𝑿𝑩 =48 6 4 0 2 0 𝒁𝒋
0 0 0 2 0 𝒁𝒋 − 𝑪𝒋
𝐙 = 48
∴ Max z=48
𝒙𝟏 = 𝟖, 𝒙𝟐 = 𝟎𝟓
Exercise:
1. Solve the following LPP by Big-M method(Method of Penalities)
𝑴𝒂𝒙 𝒁 = 𝟒𝒙𝟏 + 𝟔𝒙𝟐
Subject to
𝒙𝟏 + 𝟐𝒙𝟐 ≤ 𝟐,
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
i.e, 𝑍 ∗ = 0. 𝑥1 + 0. 𝑥2 + − − − − − − −𝑎1 − 𝑎2
i)𝑀𝑎𝑥𝑍 ∗ < 0 and artificial variable at the level in this case given LPP has no solution.
ii) 𝑀𝑎𝑥𝑍 ∗ = 0 and no artificial variable appears in the basis then proceed to phase-II.
iii)𝑀𝑎𝑥𝑍 ∗ = 0 and artificial variable appears in the basis at zero level then proceed to phase-II.
Phase-II: Assign original cost to variables in the objective function and solve by simplex method
for this phase. We use the table obtained at the end of the phase-I.
Solved problems
Subject to
𝒙𝟏 + 𝟐𝒙𝟐 ≤ 𝟐,
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
𝑥1 + 2𝑥2 + 𝑥3 = 2
8x1 + 6x2 − 𝑥4 + a1 = 24
𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , a1 ≥ 0
Max Z ∗ = 0. 𝑥1 + 0. 𝑥2 + 0. 𝑥3 + 0. 𝑥4 − a1 .
𝑥3 = 2, a1 = 24
Simplex starting table is
𝑪𝒋 0 0 0 0 -1
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 a1 =
𝑿𝒌 𝑿𝟏
0 𝑥3 2 1 2 1 0 0 2←
1 3
-1 a1 24 8 6 0 -1
Z
=𝑪𝑩 𝑿𝑩 = -8 -6 0 1 -1 𝒁𝒋
-24
-8 -6 0 1 0 𝒁𝒋 − 𝑪𝒋
↑
𝑪𝒋 0 0 0 0 -1
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 a1 =
𝑿𝒌 𝑿𝟏
0 𝑥1 2 1 2 1 0 0 2←
1 3
-1 a1 8 0 -10 -8 -1
Z -1
=𝑪𝑩 𝑿𝑩 = 0 10 8 1 𝒁𝒋
-8
0 10 8 1 0 𝒁𝒋 − 𝑪𝒋
Here all 𝑍𝑗 − 𝐶𝑗 greater than or equal to zero,but an artificial variable is present in the basis(i.e
a1 ).
Subject to
𝟑𝒙𝟏 − 𝒙𝟐 − 𝒙𝟑 ≥ 𝟎,
𝒙𝟏 − 𝒙𝟐 + 𝒙𝟑 ≥ 𝟐,
𝒙𝟏 + 𝒙𝟐 ≥ 𝟑
𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 ≥ 𝟎
3𝑥1 − 𝑥2 − 𝑥3 − 𝑥4 + 𝑎1 = 3
x1 − x2 − x3 − 𝑥5 + a2 = 2
𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , x5 , a1 , a2 ≥ 0
Max Z ∗ = 0. 𝑥1 + 0. 𝑥2 + 0. 𝑥3 + 0. 𝑥4 + 0. 𝑥5 − a1 − a2
a1 = 3, a2 = 2
𝑪𝒋 0 0 0 0 0 -1 -1
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 a1 a2
=
𝑿𝒌 𝑿𝟏
-1 a1 3 3 -1 -1 -1 0 1 0 1←
-1 a2 2 1 -1 1 0 -1 0 1 2
Z
=𝑪𝑩 𝑿𝑩 =
-4 2 0 1 1 -1 -1 𝒁𝒋
-5
-4 2 0 1 1 0 0 𝒁𝒋 − 𝑪𝒋
↑
First iteration table is
𝑪𝒋 0 0 0 0 0 -1
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 a2
=
𝑿𝒌 𝑿𝟑
0 𝒙𝟏 1 1 -1/3 -1/3 -1/3 0 0 ---
-1 a2 1 0 -2/3 4/3 1/3 -1 1 𝟑/𝟒 ←
Z
=𝑪𝑩 𝑿𝑩 =.
0 2/3 -4/3 -1/3 1 -1 𝒁𝒋
-1
𝑪𝒋 0 0 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 =
𝑿𝒌 𝑿𝟑
0 𝒙𝟏 5/4 1 -1/2 0 -1/4 -1/4
Z
=𝑪𝑩 𝑿𝑩 =.
0 0 0 0 0 𝒁𝒋
-1
0 0 0 0 0 𝒁𝒋 − 𝑪𝒋
Here all 𝑍𝑗 − 𝐶𝑗 greater than or equal to zero and no artificial variable in the basis.
Max Z ∗ = 0
Phase-II
New objective function is
15
Max Z ∗ = −( 2 ). 𝑥1 + 3. 𝑥2 + 0. 𝑥3 + 0. 𝑥4 + 0. 𝑥5
𝑪𝒋 -(15/2) 3 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 =
𝑿𝒌 𝑿𝟑
-15/2 𝒙𝟏 5/4 1 -1/2 0 -1/4 -1/4
Z
=𝑪𝑩 𝑿𝑩 =.
-(15/2) 15/4 0 15/8 15/8 𝒁𝒋
-(75/8)
Here all 𝑍𝑗 − 𝐶𝑗 ≥ 0
75
Max Z ∗ = −( )
8
75
Min Z= 8
𝟐
𝒙𝟏 = 𝟒 , 𝒙𝟐 = 𝟎
Exercise:
Subject to
𝟒 + 𝟔𝒙𝟐 + 𝟑𝒙𝟑 ≤ 𝟎,
𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 ≥ 𝟎
Special cases of Simplex method
Degeneracy:
At the stage of improving the solution during simplex procedure minimum ratio is
determined in the last column of simplex table to find the leaving variable. If the minimum ratio
is same then the problem is said to be degenerate. The following procedure is to resolve
degeneracy.
ii) Re arrange the columns forming identity matrix comes in first order.
𝐸𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑜𝑓 1𝑠𝑡 𝑐𝑜𝑙𝑢𝑚𝑛 𝑜𝑓 𝑢𝑛𝑖𝑡 𝑚𝑎𝑡𝑟𝑖𝑥
iii) Find the minimum ratio {𝑐𝑜𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑒𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑜𝑓 𝑒𝑛𝑡𝑒𝑟𝑖𝑛𝑔 𝑐𝑜𝑙𝑢𝑚𝑛 }
iv) If the minimum ratio is also same then find the ratio of
𝐸𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑜𝑓 2𝑛𝑑 𝑐𝑜𝑙𝑢𝑚𝑛 𝑜𝑓 𝑢𝑛𝑖𝑡 𝑚𝑎𝑡𝑟𝑖𝑥
{𝑐𝑜𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑖𝑛𝑔 𝑒𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑜𝑓 𝑒𝑛𝑡𝑒𝑟𝑖𝑛𝑔 𝑐𝑜𝑙𝑢𝑚𝑛 } , if the ratio is again same, repeat the above procedure
till we get unique minimum ratio.
Note: There is no need to apply the procedure when a tie for artificial variable and other
variables occurred.
Subject to
𝟒𝒙𝟏 + 𝟑𝒙𝟐 ≤ 𝟖,
𝟒𝒙𝟏 − 𝒙𝟐 ≤ 𝟖,
𝒙𝟏 , 𝒙𝟐 ≥ 𝟎
4𝑥1 + 3𝑥2 + 𝑥3 = 12
4x1 + x2 + 𝑥4 = 8
𝟒𝒙𝟏 − 𝒙𝟐 + 𝑥5 = 𝟖
𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , x5 ≥ 0
Max z = 2𝑥1 + 𝑥2 + 0. 𝑥3 + 0. 𝑥4 − 0. x5 .
𝑥3 = 12, x4 = 8, 𝑥5 = 8
𝑪𝒋 2 1 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 x5 =
𝑿𝒌 𝑿𝟏
0 𝑥3 12 4 3 1 0 0 𝟑
0 2
0 x4 8 4 1 0 1
1 2
0 𝑥5 8 4 -1 0 0
Z 0
0 =𝑪𝑩 𝑿𝑩 = 0 0 0 0 𝒁𝒋
0
-2 -1 0 0 0 𝒁𝒋 − 𝑪𝒋
↑
First iteration table is
𝑪𝒋 2 1 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟑 𝒙𝟒 x5 𝒙𝟏 𝒙𝟐 =
𝑿𝒌 𝑿𝟏
0 𝑥3 12 1 0 0 4 3 --
1/4
0 x4 8 0 1 0 4 1
0←
0 𝑥5 8 0 0 1 4 -1
Z 0
=𝑪𝑩 𝑿𝑩 = 0 0 0 0 𝒁𝒋
0
0 0 0 -2 -1 𝒁𝒋 − 𝑪𝒋
↑
𝑪𝒋 2 1 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟑 𝒙𝟒 x5 𝒙𝟏 𝒙𝟐 𝒙𝒌
=
𝒙𝟓
0 𝑥3 4 1 0 -1 0 4 1
0←
1 x2 0 0 1 -1 0 2
---
2 𝑥1 2 0 0 1/4 1 -1/4
Z -1/2
=𝑪𝑩 𝑿𝑩 = 0 0 1/2 2 𝒁𝒋
4
0 0 1/2 0 -3/2 𝒁𝒋 − 𝑪𝒋
↑
Third iteration table is
𝑪𝒋 2 1 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟑 𝒙𝟒 x5 𝒙𝟏 𝒙𝟐 =
𝑿𝒌 𝑿𝟏
4
0 𝑥3 4 1 -2 1 0 0
←
---
1 x2 0 0 1/2 -1/2 0 1
16
2 𝑥1 2 0 1/8 1/8 1 0
Z 1
=𝑪𝑩 𝑿𝑩 = 0 3/4 -1/4 2 𝒁𝒋
4
0 3/4 -1/4 0 0 𝒁𝒋 − 𝑪𝒋
↑
𝑪𝒋 2 1 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟑 𝒙𝟒 x5 𝒙𝟏 𝒙𝟐 =
𝑿𝒌 𝑿𝟏
0 𝑥5 4 1 -2 1 0 0
1 x2 2 1/2 -1/2 0 0 1
Here all 𝑍𝑗 − 𝐶𝑗 ≥ 0
Max Z = 5
𝟑
𝒙𝟏 = 𝟐 , 𝒙𝟐 = 𝟐
Unbounded Solution
Subject to
𝟑𝒙𝟏 − 𝒙𝟐 + 𝒙𝟑 ≤ 𝟏𝟎,
𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 ≥ 𝟎
3x1 − x2 + x3 + x5 = 12
𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , x5 , x6 ≥ 0
𝑥4 = 2, 𝑥5 = 10, 𝑥6 = 0
Simplex starting table is
𝑪𝒋 2 3 4 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 𝑥6 =
𝑿𝒌 𝑿𝟑
0 𝒙𝟒 2 -1 -5 -9 1 0 0 ------
1 0 10
0 𝑥5 10 3 -1 1 0
0 1 -------
0 𝑥6 2 2 3 -7 0
𝑍 0 0
0 0 0 0 𝒁𝒋
=𝑪𝑩 𝑿𝑩 =0
-2 -3 -4 0 0 0 𝒁𝒋 − 𝑪𝒋
↑
First iteration table is
𝑪𝒋 2 3 4 0 0 0
𝑿𝑩 𝑿𝑩
𝑪𝑩 𝑩𝑽 𝑿𝑩 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝑥5 𝑥6 =
𝑿𝒌 𝑿𝟑
0 𝒙𝟒 92 26 -14 0 1 9 0 ------
4 𝑥3 10 3 -1 1 0
1 0 ------
7 1 -------
0 𝑥6 72 23 -4 0 0
𝑍 4 0
12 -4 4 0 𝒁𝒋
=𝑪𝑩 𝑿𝑩 =40
10 -7 0 0 4 0 𝒁𝒋 − 𝑪𝒋
↑
Since all elements in column of pivot element are negative. Is is unbounded solution.
UNIT-II
Transportation Problem
The transportation problem is a particular case of LPP. In which the objective is to transport
various quantities of single homogeneous commodity that are initially stored at various origins in
such a way that the total transportation cost is minimum.
In a Transportation Problem the total availability should be equal to total demand. Otherwise
it is said to be unbalanced transportation problem.
Ex: A Soft drink manufacturing company has m plants situated in different cities. The total
manufacture product is observed by n retail shops located in different cities. We have to
determine the transportation schedule that minimizes the total cost of soft drink from various
plants to various retail shops.
Let 𝐶𝑖𝑗 represents unit transportation cost for transporting the unit from origin I to destination j.
The objective is to determine the number of units to be transported from origin i to destination j
so that total transportation cost is minimum.
If 𝑥𝑖𝑗 is the number of units transported from origin i to destination j , then equivalent LPP
will be
𝑍 = ∑𝑚 𝑛
𝑖=1 ∑𝑗=1 𝑥𝑖𝑗 𝑐𝑖𝑗
S to C: ∑𝑛𝑗=1 𝑥𝑖𝑗 = 𝑎𝑖
∑𝑚
𝑖=1 𝑥𝑖𝑗 = 𝑏𝑗
And 𝑥𝑖𝑗 ≥ 0
Note : 1) If ∑𝑚 𝑛
𝑖=1 𝑎𝑖 = ∑𝑗=1 𝑏𝑗 then the transportation problem is said to balance
Transportation Problem.
2) If ∑𝑚 𝑛
𝑖=1 𝑎𝑖 ≠ ∑𝑗=1 𝑏𝑗 then the transportation problem is said to unbalanced Transportation
Problem.
Finding IBFS
Procedure:
Step-1: The first assignment made in the cell occupying the upper left hand (North-West Corner)
in the table. The minimum possible amount is to be allocated as
𝑥11 = 𝑚𝑖𝑛(𝑎1 , 𝑏1 ).
This value of 𝑥11 is then entered in the cell (1, 1) of the Transportation table.
Step-2: If 𝑏1 > 𝑎1 move vertically down towards the second row and make the second
allocation of amount 𝑥12 = 𝑚𝑖𝑛(𝑎1 − 𝑥11 , 𝑏2 ) in the cell (1, 2).
If 𝑏1 < 𝑎1 move horizontally right side to the Second column and make the second
allocation of amount 𝑥21 = 𝑚𝑖𝑛(𝑎2 , 𝑏1 − 𝑥11 ) in the cell (2, 1).
If 𝑏1 = 𝑎1 then there is a tie for second element, we can make the allocation randomly.
Step-3: Start from the North-West Corner after Transportation table and repeat step-1 and step-2
until all requirements are satisfied.
Solved problems
1. Find IBFS by North-West Corner Rule
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 6 4 1 5 14
𝑶𝟐 8 9 2 7 16
𝑶𝟑 4 3 6 2 5
Requirement 6 10 15 4
Solution.
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (6) 14 8
6 4 1 5
𝑶𝟐 16
8 9 2 7
𝑶𝟑 5
4 3 6 2
Requirement 6 10 15 4
0
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (6) (8) 14 8 0
1 5
6 4
𝑶𝟐 8 9 2 7 16
𝑶𝟑 4 3 6 2 5
Requirement 6 10
15 4
0 2
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (6) (8) 14 8 0
6 4 1 5
𝑶𝟐 (2) 16 14
8 2 7
9
𝑶𝟑 4 3 6 2 5
Requirement 10
6
2 15 4
0
0
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (6) (8) 14 8 0
1 5
6 4
𝑶𝟐 (2) (14) 16 14 0
8 7
9 2
𝑶𝟑 (1) (4) 5 4 0
4 3
6 2
Requirement 10 15
6 4
2 1
0 0
0 0
Procedure:
Step-2: If 𝑥𝑖𝑗 = 𝑏𝑗 then cross out j th column of the Transportation table and decrease 𝑎𝑖 𝑎𝑛𝑑 𝑏𝑗
then go to step-3.
Step-3: If 𝑥𝑖𝑗 = 𝑎𝑖 then cross out i th row of the Transportation table and decrease 𝑏𝑗 𝑎𝑛𝑑 𝑎𝑖
then go to step-4.
Step-4: Repeat step 2 and Step 3 for resulting reduced Transportation table until all the
requirements are satisfied. Whenever the minimum cost is not unique make an arbitrary choice
(select randomly).
Solved problems
1. Find IBFS by Matrix Minima Method
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 6 4 1 5 14
𝑶𝟐 8 9 2 7 16
𝑶𝟑 4 3 6 2 5
Requirement 6 10 15 4
Solution.
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (14) 14 0
4 5
6 1
𝑶𝟐 8 9 2 7 16
𝑶𝟑 4 3 6 2 5
Requirement 6 10 15 4
1
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (14) 14 0
4 5
6 1
𝑶𝟐 (1) 16 15
8 9 7
2
𝑶𝟑 4 3 6 2 5
Requirement 15
6
10 1 4
0
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (14) 14 0
4 5
6 1
𝑶𝟐 (1) 16 15
8 9
2 7
𝑶𝟑 (4) 5 1
4 3 6
2
Requirement 15
6 4
10 1
0
0
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (14) 14 0
4 5
6 1
𝑶𝟐 (6) (1) 16 15 9
9
8 2 7
𝑶𝟑 (1) (4) 5 1 0
4 6
3 2
Requirement 6 15
10 4
0 1
9 0
0
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (14) 14 0
6 4 1 5
𝑶𝟐 (6) (9) (1) 16 15 9 0
8 9 2 7
𝑶𝟑 (1) (4) 5 1 0
4 6
3 2
Requirement 6 10 15
4
0 9 1
0
0 0
Step-1: For each Transportation table identify smallest and next to smallest cost. Determine the
difference between them for each row and column. These are called penalties. Put them along the
side of the Transportation table by enclosing them in the parenthesis against the rows and
columns.
Step-2: Identify rows or column with largest penalty among all rows and column. If a tie occurs
use any arbitrary choice. Let the Largest penalty corresponding to i th row then 𝐶𝑖𝑗 is the
smallest cost in the i th row. Allocate the amount 𝑥𝑖𝑗 = 𝑚𝑖𝑛(𝑎𝑖 , 𝑏𝑗 ) in the cell (𝑖 , 𝑗) and cross
out the ith row and jth column in usual manner.
Step-3:Again compute the column and row penalties for the reduced Transportation table and
then go to step-2.Repeate the procedure until all the requirements are satisfied.VAM is also
known as penalty method or method of penalty.
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 6 4 1 5 14
𝑶𝟐 8 9 2 7 16
𝑶𝟑 4 3 6 2 5
Requirement 6 10 15 4
Solution.
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 6 4 1 5 14 (3)
𝑶𝟐 (15) 16 (5) ←
8 9 7
2 1
𝑶𝟑 4 3 6 2 5 (1)
Requirement 6 10 15 0 4
(2) (1) (1) (3)
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 6 4 1 5 14 (3)(1)
𝑶𝟐 (15) 16 (5) ←(1)
8 9 2 7 1
𝑶𝟑 (4) 5 (1) (1)
4 3 6
2 1
Requirement 6 10 15 0 4 0
(2) (1) (1) (3)
(2) (1) (3)
↑
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 6 4 1 5 14 (3)(1)(2)
𝑶𝟐 (15) 16 (5) ←(1)(1)
8 9 2 7 1
𝑶𝟑 (1) (4) 5 (1) (1)(1)
3 6
4 2 1 0
Requirement 6 5 10 15 0 4 0
(2) (1) (1) (3)
(2) (1) (3)
(2) (1)
↑
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (10) 1 5 14 (3)(1)(2)(2)
6
4 4
𝑶𝟐 (15) 16
8 9 2 7 (5) ←(1)(1)(1)
1
𝑶𝟑 (1) (4) 5 (1) (1)(1)
3 6
4 2 1 0
Requirement 6 5 10 0 15 0 4 0
(2) (1) (1) (3)
(2) (1) (3)
(2) (1)
↑ (5)
(2) ↑
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑶𝟏 (4) (10) 1 5 14 (3)(1)(2)(2)
6 4 4 0
𝑶𝟐 (15) 16
(1)
9 2 7 (5) ←(1)(1)(1)
8
1 0
𝑶𝟑 (1) (4) 5 (1) (1)(1)
3 6
4 2 1 0
Requirement 6 5 0 10 0 15 0 4 0
(2) (1) (1) (3)
(2) (1) (3)
(2) (1)
↑ (5)
2) ↑
Total cost: 24+40+8+30+4+8=114
Step-4: Find IBFS by Vogel’s approximation method or by any of the given methods.
Step-6: For a non-degenerate transportation problem for all basic cells (the cells which have
allocations) Solve the system of Equations 𝑢𝑖 + 𝑣𝑗 = 𝐶𝑖𝑗 for which the cell ( i , j) is in the basis
starting initially with 𝑢𝑖 = 0 or 𝑣𝑗 = 0.(Generally we will take for the row or column which
contains max number of allocations. enter the values of 𝑢𝑖 & 𝑣𝑗 on the transportation table.
Step-7: To test whether the IBFS obtained in step-4 is optimum or not. Compute the difference
𝑑𝑖𝑗 = 𝐶𝑖𝑗 − (𝑢𝑖 + 𝑣𝑗 ) for all non basic cells.(not allocated cell)
Step-9: If at least one 𝑑𝑖𝑗 < 0 , select variable 𝑥𝑟𝑠 (having the most negative 𝑑𝑟𝑠 ) to enter the
basis.
Step-10: Let the variable 𝑥𝑟𝑠 enter the basis, allocate an unknown quantity 𝜃 to the cell ( r , s),
then construct a loop that starts and ends at the cell (r, s) and connects some of the basic cells.
The amount 𝜃 is added and subtracted from the transition cells of the loop in such a manner that
availabilities and requirements are satisfied.
Step-11: Assign a largest possible value 𝜃 in such a way that the value of at least one basic
variable becomes zero and other basic variables remain non-negative.
The basic cell whose allocation has been made zero will leave the basis.
Step-12: Now return to step-6 until optimum solution has been obtained.
Solved problems:
1. Find optimum solution for following transportation problem
𝑫𝟏 𝑫𝟐 𝑫𝟑 𝑫𝟒 Availability
𝑺𝟏 19 30 50 10 7
𝑺𝟐 70 30 40 60 9
𝑺𝟑 40 8 70 20 18
Requirement 5 8 7 14
Solution. First we find Initial Basic Feasible Solution by Vogel’s Approximation Method
(5) (2)
30 50
19 10
(7) (2)
70 30
40 60
(8) (10)
40 70
8 20
Now we solve the system of equations 𝑐𝑖𝑗 = 𝑢𝑖 + 𝑣𝑗 and then find 𝑑𝑖𝑗 = 𝑐𝑖𝑗 − (𝑢𝑖 + 𝑣𝑗 ) for non
basic cells.
19 -2 -10 10 𝑢1 = 10
69 48 40 60 𝑢2 = 60
29 8 0 20 𝑢3 = 20
𝑣1 = 9 𝑣2 = −12 𝑣3 = −20 𝑣4 = 0
. 30 50 . . -2 -10 .
70 30 . . - 69 48 . .
40 . 70 . 29 . 0 . -
. 32 60 .
= 1 -18 . .
11 . 70 .
Here one element of 𝑑𝑖𝑗 is less than zero. Form a loop starting at the cell (2,2) and end at the
same cell by connecting some of the basic cells.
The amount ‘θ’ to be added and subtracted from the transition cells of the loops in such a way
that the corners (availabilities) and requirements satisfied.
. 32 60 .
∴
+θ 2- θ
-18 .
1 .
.
. 10+ θ
11 8- θ 70
Θ=Min{ 2- θ,8- θ}
(5) (2)
30 50
19 10
(2) (7) (0)
70
30 40 60
(6) (12)
40 70
8 20
19 -2 8 10 𝑢1 = −32
51 30 40 42 𝑢2 = 0
29 8 18 20 𝑢3 = −22
𝑣1 = 51 𝑣2= 30 𝑣3 = 40 𝑣4 = 42
. 30 50 . . -2 8 .
70 . . 60 - 51 . . 42
40 . 70 . 29 . 18 . -
. 32 42 .
19 . . 18
11 . 52 .
=
Here elements of 𝑑𝑖𝑗 are greater than or equals to zero. Therefore the optimal solution has been
obtained.
∴ Total cost=743
Case (i): While finding the IBFS allocate a small quantity to one or more basic non basic cells,
so that number of basic cells becomes (m+n-1). This quantity is denoted by ∆ . This quantity is
introduced to least cost independent cell which does not form a loop.
Case (ii): During the testing of optimal solution to resolve degeneracy which occurs during
finding optimal solution. The quantity may be allocated to one or more basic cell which has
become non basic recently to have (m+n-1) in the new solution.
𝑫𝟏 𝑫𝟐 𝑫𝟑 Availability
𝑶𝟏 8 7 3 60
𝑶𝟐 3 8 9 70
𝑶𝟑 11 3 5 80
Requirement 50 80 80
Solution.
(60)
7
8 3
(20)
3 8
9
(50) (80) 5
11 3
Here m+n-1=3+3+-1=5
To resolve degeneracy we can assign ∆ to cells (1,2), (2,2), (3,1), (3,3) which does not form
loop. But we assign to (3,3) because it is having least cost i.e 5.
(60)
7
8 3
(20)
3 8
9
(50) (80) (∆ )
11 3 5
-3 1 3 𝑢1 = 3
8 7 . 3 7 9 𝑢2 = 9
. 8 . 𝑢3 = 5
- -1 3 5
11 . . 𝑣1 = −6 𝑣2= − 2 𝑣3 = 0
11 6 .
= . 1 .
12 . .
Here elements of 𝑑𝑖𝑗 are greater than or equals to zero. Therefore the optimal solution has been
obtained.
𝑶𝟏 → 𝑫𝟑
𝑶𝟐 → 𝑫𝟏
𝑶𝟐 → 𝑫𝟑
𝑶𝟑 → 𝑫𝟐
∴ Total cost=180+150+180+240=750
𝑫𝟏 𝑫𝟐 𝑫𝟑 Availability
𝑶𝟏 25 17 25 300
𝑶𝟐 15 10 18 500
Requirement 300 300 500
𝑫𝟏 𝑫𝟐 𝑫𝟑 Availability
𝑶𝟏 25 17 25 300
𝑶𝟐 15 10 18 500
𝑶𝟑 0 0 0 300
Requirement 300 300 500
(100) (200)
25 17 25
(300) (200)
15 10 18
0 (300)
0
0
Here m+n-1=3+3+-1=5
22 17 25 𝑢1 = −7
25 . . 15 10 18 𝑢2 = 0
. . 18 𝑢3 = −18
- -3 -8 0
0 0 . 𝑣1 = 15 𝑣2 = 10 𝑣3 = 18
3 . .
= . . 0
3 8 .
Here elements of 𝑑𝑖𝑗 are greater than or equals to zero. Therefore the optimal solution has been
obtained.
𝑶𝟏 → 𝑫𝟐 =100× 𝟏𝟕 = 𝟏𝟕𝟎𝟎
𝑶𝟏 → 𝑫𝟑 =200× 𝟐𝟓 = 𝟓𝟎𝟎𝟎
𝑶𝟐 → 𝑫𝟏 =300× 𝟏𝟓 = 𝟒𝟓𝟎𝟎
𝑶𝟐 → 𝑫𝟐 =2000
𝑶𝟑 → 𝑫𝟑 =0
∴ Total cost=13200
Assignment Problem
It is an optimum technique whose aim is to assign number of jobs equal to number of persons
at a minimum cost or a maximum profit. Suppose there are n jobs and n persons are available for
doing this job. Assume that each person can do each job with varying degree of efficiency. If 𝐶𝑖𝑗
be the cost and if i th person is assigned to j th job, the problem is to find an assignment so that
the total cost for performing all jobs is minimum. The general assignment problem as follows.
Jobs
1 2 ---------- J ---------- 𝑛
This method is used to assign men to offices, machines to jobs, drivers to trucks, trucks to
delivery routes etc.
To solve an Assignment Problem using this method follow the following steps.
Step-2: Test whether it is balanced or not, if it unbalanced make it balanced by adding dummy
rows or dummy columns.
Step-4: Locate the smallest element in each row and subtract it from all the elements of
corresponding row.
Step-5: Locate the smallest element in each column of the matrix obtained in above step and
subtract it from all the elements of corresponding column.
Step-6: Draw the minimum number of horizontal and vertical lines to cover all zeros.
Step-7: Test whether the number of lines drawn be equal to number of rows.
Step-8: If number of lines drawn equal to number of rows, it is possible to assign, otherwise go
to next step.
Step-9: Locate the smallest uncovered elements by the lines and subtract it from all uncovered
elements and add it at the intersection of lines.
Step-10: Repeat the steps 6 to 9 until number of lines drawn is equal to number of rows.
Step-11: Locate the rows and columns which is having a single zero and assign that element and
delete other zeros in the corresponding row or column. If no row or column having a single zero
then assign any zero and delete other zeros in the corresponding row or column. Repeat this
procedure until all elements are satisfied.
1. Following table gives man hours to do tasks by four persons. Find the optimum
assignment to minimize the total man hours.
I II III IV
A 9 27 20 12
B 14 29 8 27
C 39 20 19 16
D 20 28 26 10
Solution.
I II III IV Column
Minimum
A 9 27 20 12 9
B 14 29 8 27 8
C 39 20 19 16 16
D 20 28 26 10 10
I II III IV
A 0 18 11 3
B 6 21 0 19
C 23 4 3 0
D 10 18 16 0
Row 0 4 0 0
minimum
I II III IV
A 0 14 11 3
B 6 17 0 19
C 23 0 3 0
D 10 14 16 0
4
3
2
B→III=8
C→II=20
D→IV=10
2. In a textile emporium there are 4 sales men available to handle any counter and service time are
given in below table. How should the sales men assigned to the counters so as to minimize the total
service time.
I II III IV
A 41 72 39 52
B 22 29 49 65
C 27 39 60 51
D 45 50 48 52
Solution.
I II III IV Column
Minimum
A 41 72 39 52 39
B 22 29 49 65 22
C 27 39 60 51 27
D 45 50 48 52 45
I II III IV
A 2 33 0 13
B 0 7 27 43
C 0 12 33 24
D 0 5 3 7
Row 0 5 0 7
minimum
I II III IV
A 2 28 0 6
B 0 2 27 36
C 0 7 33 17
D 0 0 3 0 2
1
1 3
Here number of lines=3 1
Number of rows=4
I II III IV
A 2 26 0 4 4
B 0 0 27 34 3
C 0 5 33 15
D 2 0 5 0 2
1
1
1 of rows=4
Here number of lines=Number
It is possible to assign
I II III IV
A 2 26 0 4
B 0 0 27 34
C 0 5 33 15
D 2 0 5 0
The assignment is A→ 𝐼𝐼𝐼=39
B→ 𝐼𝐼=29
C→I=27
D→IV=52
Solved Problem.
1. A company has a team of 4 sales men. And there are 4 districts; the company wants to
start its business. The capabilities of salesmen and nature of districts are vary. The company
estimates the profit per day in rupees for each sales man in each district as follows. Find the
optimum assignment and total profit.
I II III IV
A 41 72 39 52
B 22 29 49 65
C 27 39 60 51
D 45 50 48 52
Solution.
I II III IV
A 6 10 14 11
B 14 11 15 15
C 15 15 13 12
D 13 12 14 15
Maximum is 15, we subtract all the elements from this.
I II III IV Column
minimum
A 9 5 1 4 1
B 1 4 0 0 0
C 0 0 2 3 0
D 2 3 1 0 0
I II III IV
A 8 4 0 3
B 1 4 0 0
C 0 0 2 3
D 2 3 1 0 3
1
2 1
Here number of lines=3 1
Number of rows=4
Locate smallest element in uncovered elements i. e ‘1’ and subtract it from all the elements
and add at the intersection of lines.
I II III IV
A 7 3 0 3 4
B 0 3 0 0 2
C 0 0 3 4
D 1 2 1 0 3
1
1
1
Here number of lines=Number of rows=4
It is possible to assign
I II III IV
A 7 3 0 3
B 0 3 0 0
C 0 0 3 4
D 1 2 1 0
B→ 𝐼=14
C→ 𝐼I=15
D→IV=15
This is a special case of Assignment Problem. The problem is meant for allotment of salesmen to
different regions through different routes such that the variables involved such as distance, time, Cost
are optimized.
Suppose a Salesman wants to visit a certain number of cities allotted to him. He knows the distance
or cost or time of journey between every pair of cities usually denoted by 𝐶𝑖𝑗 . The Problem is to select a
route such that it starts from his home city and passes through each city one and only once and returns
to his home city in the shortest possible distance or cost or time.
If 𝐶𝑖𝑗 is the distance or cost or time from city I to city j and 𝑥𝑖𝑗 is equal to 1 if a salesman goes
directly from city I to city j and zero otherwise. In this case the problem is of minimization ∑𝑖 ∑𝑗 𝑥𝑖𝑗 𝐶𝑖𝑗
with an additional restriction that 𝑥𝑖𝑗 must be Chosen that no city is visited twice and it cannot go from
city i to i itself. The general Transportation Problem as follows.
𝐴1 𝐴2 −−− −−− 𝐴𝑛
𝐴1 ∞ 𝐶12 −−− −−− 𝐶1𝑛
𝐴2 𝐶21 ∞ −−− −−− 𝐶2𝑛
−−− −−− −−− −−− −−− −−−
−−− −−− −−− −−− −−− −−−
𝐴𝑛 𝐶𝑛1 𝐶𝑛2 −−− −−− ∞
Solved Problem.
1. A sales man want to visit 4 cities 1,2,3,4. He does not want to visit any city twice. The cost of going
from one city to another city in rupees is given below. Find the least cost route.
1 2 3 4
1 ∞ 30 80 50
2 40 ∞ 140 30
3 40 50 ∞ 20
4 70 80 130 ∞
Solution.
We find the optimum route for the given travelling sales man problem by Hungarian method
1 2 3 4 Column
minimum
1 ∞ 30 80 50 30
2 40 ∞ 140 30 30
3 40 50 ∞ 20 20
4 70 80 130 ∞ 70
1 2 3 4
1 ∞ 0 50 20
2 10 ∞ 110 0
3 20 30 ∞ 0
4 0 10 60 ∞
Row 0 0 50 0
minimum
1 2 3 4
1 ∞ 0 0 20 3
2 10 ∞ 60 0
3 20 30 ∞ 0
4 0 10 10 ∞
2
3
1
Here number of lines=3 1
Number of rows=4
Locate smallest element in uncovered elements i. e ‘10’ and subtract it from all the
elements and add at the intersection of lines.
1 2 3 4
1 ∞ 0 0 30 2
2 10 ∞ 50 0
3 20 20 ∞ 0
4 0 0 0 ∞ 1
3
1
Here number of lines=3 1
Number of rows=4
Locate smallest element in uncovered elements i. e ‘10’ and subtract it from all the
elements and add at the intersection of lines.
1 2 3 4
1 ∞ 0 0 40 2
2 0 ∞ 40 0 4
3 10 10 ∞ 0 1
4 0 0 0 ∞ 1 1
Number of rows=4
It is possible to assign.
1 2 3 4
1 ∞ 0 0 40
2 0 ∞ 40 0
3 10 10 ∞ 0
4 0 0 0 ∞
SEQUENCING
The selection of appropriate order in which jobs may be done (or) waiting customers may
be served is called Sequencing.
Suppose there are n jobs (n = 1,2,3, ------- ) each of which has to be processed at a time at each
machine ( M = A,B,C,---------- ) .The order of processing each job through M machines must
required the time on each machine is known. The problem is to find a sequence of possible
sequences for processing the jobs, so that the total allotted time for all jobs will be minimum.
Mathematically
1) n jobs 2 Machines.
2) n jobs 3 Machines.
3) n jobs m Machines.
Only two machines A and B are involved, each job is processed in the order AB and
expected processing times 𝐴1 , 𝐴2 , 𝐴3 , − − − − − − −𝐴𝑛 , 𝐵1 , 𝐵2 , 𝐵3 , − − − − − − 𝐵𝑛 are
known.
The problem is to find sequence of jobs so as to minimize the total elapsed time T.
Procedure to find Sequence of jobs in case of n jobs two machines (Johnsons Algorithm)
Step-2: If there is a time any one of the smallest processing time should be selected.
i)If the minimum is 𝐴𝑘 for some k, process the k th job first of all.
ii) If the minimum is 𝐵𝑟 for some r, process the rth job last of all.
Step-3: i) If the tie for minimum of 𝐴𝑘 and 𝐵𝑟 (i.e. 𝐴𝑘 = 𝐵𝑟 ) occurs then process the k th job
first of all and r th job in the last.
ii)If the tie for minimum occurs among 𝐴𝑖 select the job corresponding to the minimum of 𝐵𝑖 s
and process it first of all.
iii) If the tie for minimum occurs among 𝐵𝑖 select the job corresponding to the minimum of 𝐴𝑖 s
and process it in the last.
Step-4: Cross out the jobs already assigned and repeat the steps 1,2,3 arranging the jobs next to
first (or) next to last until all jobs have been assigned.
1. Find the sequence of jobs that minimizes total elapsed time to compute the following jobs on
two machines.
Jobs 1 2 3 4 5 6
A 3 12 5 2 9 11
B 8 10 9 6 3 1
Solution. Given
4 6
4 1 5 6
Jobs Machine A Machine B
2 12 10
3 5 9
4 1 3 2 5 6
To calculate idle time and minimum elapsed time corresponding to the optimal sequence we use
the following table
2. Calculate the total idle time for machine A and B and total elapse time
Jobs 1 2 3 4 5
A 5 1 9 3 10
B 2 6 7 8 4
Solution. Given
2 1
2 4 1
2 4 3 5 1
To calculate idle time and minimum elapsed time corresponding to the optimal sequence we use
the following table
Transfer of jobs is not permitted. Exact or Expected Processing times are given as
The following conditions are to be considered where either one or both holds.
Procedure
We convert n jobs 3 machines problem to n jobs 2 machines problem. For this we take 2
fictitious Machines denoted by G and H and we consider
𝐺𝑖 = 𝐴𝑖 + 𝐵𝑖 And 𝐻𝑖 = 𝐵𝑖 + 𝐶𝑖
Solved problems
1. There are 5 jobs each of which must go through the machines A,B, C in order ABC and
processing times are given as
Jobs Ai Bi Ci
1 8 5 4
2 10 6 9
3 6 2 8
4 7 3 6
5 11 4 5
Determine the sequence for these 5 jobs that will minimize the total elapse time. Also find the
idle time for machines A,B,C.
Therefore we can adopt the procedure followed in ‘n’ jobs 2 machines case. We take two
fictitious machines G&H and we find processing times Gi and Hi using the formula
Gi= Ai+ Bi
Hi= Bi + Ci
Jobs Gi Hi
1 13 9
2 16 15
3 8 10
4 10 9
5 15 9
Remaining table
Jobs Gi Hi
1 13 9
2 16 15
4 10 9
5 15 9
3 4
Jobs Gi Hi
1 13 9
2 16 15
5 15 9
3 1 4
Jobs Gi Hi
2 16 15
5 15 9
3 2 5 1 4
Exercise
1. The processing time for 6 jobs and 3 machines are given below
Job 1 2 3 4 5 6
Ai 3 12 5 2 9 11
Machine Bi 8 6 4 6 3 1
Ci 13 14 9 12 8 13
Determine the optimal sequence of jobs.
The optimal solution to this problem can be found if either one or both of the condition
satisfied.
After obtaining the new processing times 𝐺𝑖 𝑎𝑛𝑑 𝐻𝑖 we use the procedure followed in case of
n jobs 2 machines.
Solved problems
1. Solve the following sequencing problem of 4 jobs on 5 machines
Machines
M1 M2 M3 M4 M5
7 5 2 3 9
Jobs 6 6 4 5 10
5 4 5 6 8
8 3 3 2 6
Condition is satisfied.
Jobs Gi Hi
A 17 19
B 21 25
C 20 23
D 16 14
A C B D
Job M1 M2 M3 M4 M5
Time Time Time Time Time Time Time Time Time Time
in out in out in out in out in out
A 0 7 7 12 12 14 14 17 17 26
C 7 12 12 16 16 21 21 27 27 35
B 12 18 18 24 24 28 28 33 35 45
D 18 26 26 29 29 32 33 35 45 51
Here the problem is minimizing the total Elapsed time T. We use Graphical Method to
solve 2 jobs M Machine Problem.
Solved problem.
1. Using graphical method determine the optimal sequence needed to process jobs 1 &2 on
5 machines A,B,C,D,E; for each machine find the job which should be done first. Also
calculate the total time needed to complete both the jobs.
Sequence A B C D E
Job 1
Time(hours) 1 2 3 5 1
Sequence A B C D E
Job 1
Time(hours) 3 4 2 1 5
Solution.
1) Draw two axis at right angles each other represents processing time on job 1 along horizontal
axis, processing time on job 2 along vertical axis, scale used must be same for both the
jobs.2)Layout the machine times for 2 jobs on the corresponding axis in the given technological
order.3) Machine A requires 1 hour for job 1 and 4 hours for job 2. A rectangle LMNP thus
constructed for machine A. Similarly for the machine BCDE.
4) Make a program by starting from origin ‘O’ and moving through the various stages of
completion (points) till the point finish is reached. Choose path consisting only of horizontal
vertical and 450 lines, a horizontal line represents work on job 1 while job 2 remains idle, A
vertical line represents work on job 2 while job 1 remains idle and a 450 line to the base
represents simultaneous work on both jobs.
5) Find the optimal path and an optimal path is one that minimizes the idle time for job 1(vertical
movement) likewise an optimum path is one that minimizes the idle time for job 2 (horizontal
line) , obviously the optimal path is one which coincides with 450 line to the maximum extent.
Both jobs cannot processed simultaneously on one machine graphically this means that diagonal
movement through shaded areas is not allowed, a good path accordingly is chooses and drawn on
graph.
6) Find the elapsed time it is obtained by adding the idle time for either job to the processing
time for job, idle time for the chosen path is formed to be 3 hours for job 1.
Similarly idle time for job 2 is 0.
Total elapsed time=12+3=15 hours
Idle time for job 1 =3 hrs
REPLACEMENT
Introduction
The study of Replacement is concerned with situations that arise when some items such as Machines,
Men, Electric light bulbs etc need replacement due to their deteriorating efficiency, failure or break down.
The deteriorating efficiency or complete breakdown may be either gradual or all of sudden.
Eg : A Machine becomes more and more expensive due to maintenance after a number of years, An
electric bulbs fails all of a sudden, pipeline may be blocked or an employee may loses his job and so on.
In all such situations there is a need to formulate a most economic replacement policy for replacing the
deteriorating units.
Following are the situations when the replacement of certain item needed to be done.
1) An old item has failed and does not work at all (or) old item is expected to fail shortly.
2) The old item has deteriorated and work badly or requires expensive maintenance.
Generally the cost maintenance and repair of certain items (equipments) increases with time
and stage may come when this cost becomes so high that it is more economical to replace the
item by a new one.
To determine the optimum Replacement age of an item whose maintenance cost increases
with time and value of money remains static during that period.
If the Equipment is used for n years, then total cost included during this period.
𝑇𝐶
Average Annual Total Cost 𝐴(𝑛) = 𝑛
𝐶−𝑆 1 𝑛
= 𝑛
+ 𝑛 ∫0 𝑓(𝑡)𝑑𝑡
The equipment should be replaced when maintenance cost is equal to the average annual
total cost.
Here the period of time is considered as fixed and n, t takes the values 1, 2, 3, -------- then
𝐶−𝑆 1
𝐴(𝑛) == + 𝑛 ∑𝑛𝑖=1 𝑓(𝑡)
𝑛
𝐴(𝑛 − 1) ≥ 𝐴(𝑛)
In this case replace the equipment at the end of n years, if the maintenance cost (n+1) th year
is more than the total average cost in the nth year and n th year maintenance cost < the previous
year’s average total cost. .
Problems
1. A firm is considering replacement of a machine whose cost is rupees 12, 200 and the
scrap value is rupees 200, the running cost (maintenance cost) in rupees are found from
experience to be as follows
Year 1 2 3 4 5 6 7 8
Running 200 500 800 1200 1800 2500 3200 4000
cost
When the machine should be replaced.
Sol. Given
Depreciation cost=12,200-200=12000
Running cost=f(n)
In order to determine the optimal time ‘n’ when the machine should be replaced we calculate an
average total cost per year during the life of machine as shown below.
From the above table we observed that the average total cost per year A(n) is minimum in the 6th
year i. e Rs. 3167.
Also average cost in 7th year is more than the cost in 7th year. Therefore the machine should
replaced after 6 years.
2) Machine A costs Rs. 9000, annual operating costs are Rs. 200 for 1st year and then
increased by Rs. 2000 every year. Determine the best age at which the machine should
replaced. If the optimum replacement policy is followed, what will be the average yearly
cost of owning and operating the machine.
Sol. Let the machine have no resale value when replaced. For this machine from the given data
the running cost was given in the following table.
Year 1 2 3 4 5 6
Running 200 2200 4200 6200 8200 10200
cost
Cost of machine ( C ) =Rs. 9000
Scrap value(S)=0
Depreciation cost=C-S=9000
Running cost=f(n)
2) The maintenance cost is included in the beginning of different time periods. Since it is
assumed that the maintenance cost increased with time and each cost is to be paid in the start of
the period, let the money carry a rate of interest or r per year. Thus a rupee invested now will be
worth (1+r) after a year, thus (1 + 𝑟)2 after two years, and so on.
In other words (1 + 𝑟)−𝑛 is the present worth of one rupee spent after n years. The quantity
(1 + 𝑟)−𝑛 is called present worth factor (p.w.f).The expression (1 + 𝑟)𝑛 is known as payment
compound amount factor (c.a.f) of one rupee spent in n years. The procedure for determining the
weighted average cost (annualized cost) summarized in the following steps.
Step-1: Find the value of the maintenance cost for each of the year’s
𝑅𝑛 = 𝑅𝑢𝑛𝑛𝑖𝑛𝑔 𝑐𝑜𝑠𝑡
Step-2: Calculate the cost plus value of the cumulative present values obtained in step-1
i.e. 𝐶 + ∑ 𝑅𝑛 𝑣 𝑛−1
Step-3: Find the cumulative present value factor for all the years’ i.e. ∑ 𝑣 𝑛−1
Step-4: Determined the average annualized W (n) by dividing the entries obtained in step-2 by
corresponding entries obtained in step-3.
𝐶+∑ 𝑅𝑛 𝑣 𝑛−1
i.e. ∑ 𝑣 𝑛−1
1) The initial cost of an item is Rs. 15,000 and maintenance cost for different years is given
below
Year 1 2 3 4 5 6 7
Running 2500 3000 4000 5000 6500 8000 10000
cost(Rs.)
What is the replacement policy to be adopted if the rate of interest is 10% and there is no
solvage value.
From above table we observe that the average annual total cost in 5th year is minimum i. e, Rs.
7609. The item should replace at end of 5th year.
Group Replacement
This policy is concerned with items that either work perfectly or partially or inefficiently or
fails completely. This situation is generally happens when the system consists of large number of
identical low cost items that are increasingly liable to failure with age. In such cases the
replacement of individual items would incurred a set of cost which is independent of number
replaced and it may be advantageous to replace all the items at a time at a fixed interval. This
policy is known as Group Replacement Policy and is very attractive particularly when
iv) Average individual replacement would be costlier than the average group replacement.
1) Individual Replacement: Under this policy an item is replaced immediately after its failure.
2) Group Replacement: under this policy a decision will be taken so as to replace all the items
irrespective of the fact that the items have failed or not failed before the optimal time.
Step-1: Find the probability of failure of each items at the end of each period.
Step-2: Find the number of replacements made at the end of each period with reference to
probability of failures at the end of each period considering its previous Replacement.
Step-3: Calculate the cost of individual replacement at the end of each period.
Step-4: Calculate the cost of group replacement at the end of each period.
Step-5: Calculate the total cost of group replacement including individual replacement by adding
step3 and step4.
Step-6: Calculate the average cost per period by dividing the result in step-5 with period number.
Step-7: Identify cost among the average cost per period and it will be the period of Group
Replacement Policy.
1) A factory has a large number of bulbs, all of which must be in working condition. The
mortality of bulbs is given below.
Week 1 2 3 4 5 6
Proportion of bulbs 0.1 0.15 0.25 0.35 0.12 0.03
failing during the
week
If a bulb fails in service it costs rupees 3.50 to replace but if all bulbs are replaced at a time
it costs rupees 1.20 each. Find the optimum group replacement policy by assuming
thousand bulbs are available in beginning.
∴ N0= 1000
=160 bulbs
N3=281
N4=427
N5=280
N6=260
From the above table we observe that the average cost in 3rd week is minimum i. e. Rs. 1031.
The bulbs should be replaced after every 3 weeks.
UNIT-4
GAME THEORY
Introduction
Marketing, Competing, Product to enemies planning for war tactics etc. In these cases the
decision taken by one decision makers influences the decision taken by other decision makes and
final results are mainly influenced by decisions of all the parties.
This theory mainly deals with competitive problems and is based on the minimax principle.
This principle implies that each competitor will act in order to minimize this maximum loss or
maximize its minimum gain.
GAME
A Game refers to a situation in which two or more players are competitive and are involved in
activities. Where the players must follow certain set of rules and at end of which each person
receives some benefit or satisfaction while other will have loss.
PLAYER
PLAY
A Play is a set of rules and regulations which are followed by the players in a Game.
STRATEGY
A Strategy is a set of rules (programs) which specifies which of the available course of action a
player should make at each play. It is rule for decision making in advance for all the place by
which the players decide which activity to adopt.
TYPES OF STRATEGIES
1. Optimal Strategy;- An Optimal Strategy is one which provides the best situation in the Game
which involves Maximum payoff to the player. Any deviation from this strategy results in a
decreased payoff for the player.
2.Pure Strategy;- A Pure Strategy is a decision role always used by a player to choose a
particular course of action which means each player knows exactly what the other is going to do.
3. Mixed Strategy;- Mixed Strategy is a selection among pure strategies with fixed probabilities.
Payoff
A Quantitative measure of satisfaction which a player gets at the end of each play is called
Payoff.
Payoff Matrix
The Payoff Matrix is a decision Analysis tool which summarizes the Games and loss of a
decision in a tabular form. So the general payoff matrix is as follows.
𝐵1 𝐵2 ------ 𝐵𝑛
𝐴1 𝑎11 𝑎12 ------ 𝑎1𝑛
𝐴2 𝑎21 𝑎22 ------ 𝑎2𝑛
------ ------ ------ ------ ------
𝐴𝑚 𝑎𝑚1 𝑎𝑚2 ------ 𝑎𝑚𝑛
Value of the Game
It is the Expected payoff of the Game when all the players of the Game adopt their optimal
strategies.
If the value of the Game is zero, i.e. neither player wins or loses. Game is said to be fair game.
If one of the players wins (other loses) or the value of the game is non zero it is said to be
unfair game.
Saddle Point
It is the value of the game when the players adopt pure strategy. It is obtained by computing
maximin value for the row player and minimax value for the column player .If this two values
coincide the value. So obtained is called saddle point.
The Game Theory was received with a great enthusiasm initially. Later it has been found with
lot of limitations. The most prominent limitations are
1. Unrealistic Assumption:- The assumption that the play known their own as well as others
payoffs in unrealistic these are only guesses.
2. Complexity with more number of players:- As the number of players increase the problem
becomes so complex that become too difficult to solve.
4. Non-dynamic nature:- In practical markets generally changes and are very frequent and fast.
The game theory gives least scope to work out the strategy.
5. No-base for calculating payoff:- There is no correct formula and scientific base for
calculating payoffs of the game.
1) Change of strategy:- A game involves game of strategy as well as a chance. The activities are
evaluated by skill is known as a game of strategy and activities are determined by chance is
known as game of chance.
2) Number of activities;- it may be finite or infinite.
3) Number of persons;- if the n number of persons playing in a game is said to be an n-person game.
4) Number of alternatives available to each person: -In a particular course of action a finite number
of activities involving finite number of alternatives, otherwise the game said to be infinite.
5) Information to the players about the past activities of the other payer: -In this information is
completely available or not available.
Problems
1.Solve the following Game.
1 2 3 4 5
I -2 0 0 5 3
II 3 2 1 2 2
III -4 -3 0 -2 6
IV 5 3 -4 2 -6
Sol: - Given Payoff matrix is
Row Min
1 2 3 4 5
I -2 0 0 5 3 -2
II 3 2 1 2 2 1
III -4 -3 0 -2 6 0
IV 5 3 -4 2 -6 -6
Col max
5 3 1 5 6
Saddle point = 1
Max {Row Min}=Min{Col Max}=1
∴ Value of the game = 1
Best Strategy for player A is = II
Best Strategy for player B is = 3
2.Find the solution to the following Game using maximin and minimax principle.
𝐵1 𝐵2 𝐵3 𝐵4 Row min
𝐴1 2 5 8 3 2
𝐴2 -1 2 4 6 -1
𝐴30 4 3 2 0
𝐴4 1 2 3 4 1
Col max 2 5 8 6
Saddle point = 2
Max {Row Min} = Min{Col Max}=2
∴ Value of the game = 2
Best Strategy for player A is = 𝐴1
Best Strategy for player B is = 𝐵1
Where 𝑎𝑖𝑗 is the pay off corresponding to the i th strategies of player A, against j th strategy of
player B
If 𝑣 = 𝑣̅ = v then it is called fair game and v is the value of the game. The corresponding
strategy for player A and B are called the optimum strategies respectively. The fundamental
theorem of game theory says that
𝑣̅ ≥ 𝑣
Game without saddle points mixed strategies; we can solve using following methods.
1. Arithmetic Method
2. Graphical method
1. Arithmetic Method
We apply this method for (2× 2) rectangular games. Let a pay off matrix be
𝐵1 𝐵2
𝐴1 𝑎11 𝑎12 𝑝1
𝐴2 𝑎21 𝑎22 𝑝2
𝑞1 𝑞2
Such that 𝑝1 + 𝑝2 = 1
𝑞1 + 𝑞2 = 1
Where 𝑝1 , 𝑝2 are probabilities associated with the strategies of player A and 𝑞1 , 𝑞2 are
probabilities associated with player B.
𝑎22 −𝑎21
We have 𝑝1 = (𝑎11 +𝑎22 )−(𝑎12 +𝑎21 )
and 𝑝2 = 1 − 𝑝1
𝑎22 −𝑎12
𝑞1 = (𝑎11 +𝑎22 )−(𝑎12 +𝑎21 )
And 𝑞2 = 1 − 𝑞1
Problems
1. for the game with the following payoff matrix. Determine the optimum strategies and value of
the game.
𝐵1 𝐵2
𝐴1 5 1
𝐴2 3 4
Sol:-
𝐵1 𝐵2 Row
min
𝐴1 5 1 1
𝐴2 3 4 3
Col 5 4
max
Max {Row Min}≠Min{Col Max}
1 4
𝑝2 = 1 − 𝑝1 = 1 − 5 = 5
𝑎22 −𝑎12 4−1 3
𝑞1 = (𝑎11 +𝑎22 )−(𝑎12 +𝑎21 )
= (5+4)−(1+3)
= 5
3 2
𝑞2 = 1 − 𝑞1 = 1 − 5 = 5
2. There are two players in a game named as matching player and non matching player. The
matching player is paid Rs 8.00 if the two coins turn both heads and Rs 1.00 if the coins turn
both tails. The non matching player is paid Rs 3.00 when the two coins do not match. Given the
choice of being the matching or non matching player, which you choose and what would be your
strategy?
𝐻 𝑇 Row
min
𝐻 8 −3 −3
𝑇 −3 1 −3
Col 8 1
max
Clearly the given problem has no saddle point.
By Arithmetic method
𝑎 −𝑎 1−(−3) 4
𝑝1 = (𝑎 +𝑎 22)−(𝑎21 +𝑎 ) = (8+1)—(−3−3) =
11 22 12 21 15
4 11
𝑝2 = 1 − 𝑝1 = 1 − 15 = 15
𝑎22 −𝑎12 1−(−3) 4
𝑞1 = (𝑎11 +𝑎22 )−(𝑎12 +𝑎21 )
= (8+1)—(−3−3)
= 15
4 11
𝑞2 = 1 − 𝑞1 = 1 − 15 = 15
2. Graphical method
𝐵1 𝐵2 ----------- 𝐵𝑛
𝐴1 𝑎11 𝑎12 ----------- 𝑎1𝑛
𝐴2 𝑎21 𝑎22 ----------- 𝑎2𝑛
Player A has two strategies 𝐴1 , 𝐴2 with probabilities 𝑃1 , 𝑃2 . Such that 𝑃1 + 𝑃2 = 1
The 𝑚 × 2 games are also treated in the same way for these 𝑚 × 2 games we will get
minimax point which will be the lowest point on the upper boundary.
Problems
1. Find value of the game and optimal strategy for following games graphically.
I II III IV
I 1 0 4 -1
II 0 1 -2 5
Sol:-
I II III IV
I 1 0 4 -1 -1
II 0 1 -2 5 -2
1 1 4 5
Clearly the given game without saddle point. We can apply Graphical method.
Player A s expected payoff against B s strategies are given by
Player A wishes to maximize his minimum expected payoff .The highest point of intersection
H on the lower boundary of A. This point H represents the maximin expected value of the game.
I II
I 1 0
II 0 1
By Arithmetic method
𝑎 −𝑎 1−0 1
𝑝1 = (𝑎 +𝑎 22)−(𝑎21 +𝑎 = (1+1)—(0+0)
=
11 22 12 21 ) 2
1 1
𝑝2 = 1 − 𝑝1 = 1 − 2 = 2
𝑎22 −𝑎12 1−0 1
𝑞1 = (𝑎11 +𝑎22 )−(𝑎12 +𝑎21 )
= (1+1)—(0+0)
= 2
1 1
𝑞2 = 1 − 𝑞1 = 1 − 2 = 2
𝐵1 𝐵2
𝐴1 1 8
𝐴2 3 5
𝐴3 11 2
𝐵1 𝐵2 Row min
𝐴1 1 8 1
𝐴2 3 5 3
𝐴3 11 2 2
Col max 11 8
Clearly the given game without saddle point. We can apply Graphical method.
𝐵1 𝐵2
𝐴1 1 8
𝐴3 11 2
By Arithmetic method
𝑎 −𝑎 2−11 9
𝑝1 = (𝑎 +𝑎 22)−(𝑎21 +𝑎 ) = (1+2)—(8+11)
=
11 22 12 21 16
9 7
𝑝2 = 1 − 𝑝1 = 1 − 16 = 16
𝑎22 −𝑎12 2−8 3
𝑞1 = (𝑎11 +𝑎22 )−(𝑎12 +𝑎21 )
= (1+2)—(8+11)
= 8
3 5
𝑞2 = 1 − 𝑞1 = 1 − 8 = 8
Dominance Property
Sometimes we may observe that one of the pure strategies of either player is always inferior
to at least one of the remaining strategy. The superior strategy is said to be dominated by inferior
strategy .By eliminating the inferior strategy one can reduce the dimension of payoff matrix in
order to solve the game.
a) If all the elements of a row say k th row are ≤ the corresponding elements of any other r th
row then we say k th row is dominated by r th row.
b) If all the elements of a column say k th column are ≥ the corresponding elements of any
other r th column then we say k th column is dominated by r th column.
c) Dominated rows or columns may be deleted to reduce the size of the matrix.
Note: - Dominance property is not always based on the superiority of pure strategy. A given
strategy can also be said to be dominate if it is inferior to average of two or more other pure
strategies.
Problems
1. Use dominance property to obtain the optimal strategies and value of the Game.
I II III IV
I 3 5 4 2
II 5 6 2 4
III 2 1 4 0
IV 3 3 5 2
Sol: - Given
III IV
I 4 2
II 2 4
IV 5 2
By Arithmetic method
𝑎22 −𝑎21 2−5 3
𝑝1 = (𝑎11 +𝑎22 )−(𝑎12 +𝑎21 )
= (2+2)−(4+5)
= 5
3 2
𝑝2 = 1 − 𝑝1 = 1 − 5 = 5
𝑎22 −𝑎12 2−4 2
𝑞1 = (𝑎11 +𝑎22 )−(𝑎12 +𝑎21 )
= (2+2)−(4+5)
= 5
2 3
𝑞2 = 1 − 𝑞1 = 1 − 5 = 5
I II III IV
I 3 2 4 0
II 3 4 2 4
III 4 2 4 0
IV 0 4 0 8
Sol: - Given
I II III IV
II 3 4 2 4
III 4 2 4 0
IV 0 4 0 8
Column1 ≥ Column3.then delete column1.
II III IV
II 4 2 4
III 2 4 0
IV 4 0 8
Dominance property is not always based on superiority of pure strategies. A given strategy can
also said to be dominated if it is inferior to average of two or more strategies.
Column2 ≥ Average of column 3 and column4.Then delete column2.
III IV
II 2 4
III 4 0
IV 0 8
Row2 ≤ Average of row 3 and row 4.Then delete row2.
III IV
III 4 0
IV 0 8
By Arithmetic method
𝑎 −𝑎 8−0 2
𝑝1 = (𝑎 +𝑎 22)−(𝑎21 +𝑎 = (4+8)−(0+0)
=
11 22 12 21 ) 3
2 1
𝑝2 = 1 − 𝑝1 = 1 − 3 = 3
𝑎22 −𝑎12 8−0 2
𝑞1 = (𝑎11 +𝑎22 )−(𝑎12 +𝑎21 )
= (4+8)−(0+0)
= 3
2 1
𝑞2 = 1 − 𝑞1 = 1 − 3 = 3
INVENTORY MODELS
The word inventory means simply a stock of ideal resources of any kind having an economic
value. Inventory means a physical stock of goods which is kept in hand for smooth and efficient
running of an organization it may consists raw materials, spare parts, finished goods, human
resources, financial resources etc.
The problems faced by different organizations have necessitated the use of scientific
techniques in the management of inventories known as inventory control or inventory
management.
The Inventory Cost
1. Set up cost: - The cost associated with setting of the machinery before starting the production
and is independent of the quantity ordered for production.
2. Ordering Cost: - This cost associated with ordering of raw material for production.
Advertisement, consumption of stationary and postage, telephone charges, rent for space used by
purchase department, travelling expenditure etc. constitute the ordering cost.
3. Purchase or Production cost: - This is cost of purchasing or producing a unit of item. This is
very important when discounts are allowed.
4. Carrying or holding cost: - The cost associated with the storage costs like rent, interest on
the money locked up , insurance of stored equipment ,production, taxes, depreciation of
equipment etc.
5. Shortage or stock out cost: - The penalty cost for running out of stock .This cost includes the
loss of potential profit through sales of items and loss of good will, in terms of permanent loss of
customers and its associated loss profit in future sales.
6. Salvage cost or selling price: - When demand for certain commodity is affected by quality
stocked, decision problem based on profit maximization criterion includes the revenue from
selling.
Economic order quantity: The quantity to be ordered in the most economical way is called
Economic Order Quantity (EOQ).
(i) Demand: The nature of demand must be known for inventory control. Demand
may be deterministic or Probabilistic.
(ii) Order cycle: The period of time between two consecutive placements of orders is
called order cycle.
(iii) Lead Time: When an order is placed the supplier takes some time to supply the
item. Lead time is the time to between the placement of an order and receipt of
materials.
(iv) Replenishment of Stock: Replenishment of stock may occur instantaneously
without lead time or uniform replenishment may occur when the goods are
manufactured by the factory itself.
(v) Planning horizon(Period): The duration of time over which a particular inventory
level will be maintained is called planning horizon.
Inventory Models:-
Model 1: Demand rate uniform and production rate is infinite(or) Economic Order
quantity model with uniform demand (or) Wilson Harris Simple EOQ:
Cc= Carrying cost (or) Holding cost per unit per year
P= Price of item
Total annual order cost=Order cost per order× No. of orders per year
𝐷
i.e AOC= C0 × 𝑄
2𝐶0 𝐷
Q0=√ 𝐶𝑐
Note: If carrying cost or Holding cost is expressed in terms of the value of stock held then
Cc=i.p
𝟐𝑪𝟎 𝑫
Q0=√ Here ‘i’ is interest
𝒊.𝒑
Problems:
1. A company uses 10000 units per year of an item. The purchase price is Rs. 25 per
order. The interest on carrying cost per year is 12% of the inventory value and cost of
item is Rs.1. Then find the EOQ, the no. of orders per year, if the lead time is 4 weeks
and assuming 50 working weeks per year find re order point.
Sol. Given
𝟐𝑪𝟎 𝑫
(i) EOQ= √ 𝒊𝑷
𝟐×𝟐𝟓×𝟏𝟎𝟎𝟎𝟎
=√ 𝟎.𝟏𝟐×𝟏
=2041.24≅ 2041
𝐷 10000
(ii) The no. of orders per year = 𝑄= 2041 =4.899 ≅ 5
50
(iii) 𝑤𝑒𝑒𝑘𝑠= 10 weeks
5
10-4 = 6 weeks
𝑁𝑜.𝑜𝑓 𝑤𝑜𝑟𝑘𝑖𝑛𝑔 𝑤𝑒𝑒𝑘𝑠 𝑝𝑒𝑟 𝑦𝑒𝑎𝑟
Re order period= 𝑁𝑜.𝑜𝑓 𝑜𝑟𝑑𝑒𝑟𝑠 𝑝𝑒𝑟 𝑦𝑒𝑎𝑟
=10 weeks
2. A company uses 24000 units of raw material which costs rupees 12.5 per unit. Placing
each order costs Rs.22.5 and carrying cost is 5.4% per year of the average inventory.
Find the economic order quantity including the cost of material.
𝟐𝑪𝟎 𝑫
(i) EOQ= √ 𝒊𝑷
1264.91≅ 𝟏𝟐𝟔𝟓
Cc =i ×P
300853.15
Model -2: Demand rate not uniform and Production rate infinite:
In this model assumptions are same as model-1with exception that in place of uniform
demand. We were given total demand. And since annual ordering cost equal to annual
carrying cost will have same formulae as in model-1.
Model -3:
This model assumes to replenish continuously till the maximum quantity is reached due to
the fact that in many situations the amount ordered will not be receivable all at once. But will
be available at finite supply rate. The following points are assumed in this model.
i.e r >d
𝟐𝑪𝟎 𝑫
(ii) 1) EOQ= √ 𝒊𝑷
2𝐶0 𝐷 𝑟
1) Economic batch quantity(EBQ)= 𝑄𝑜𝑝𝑡 = √ (𝑟−𝑑)
𝐶𝑐
𝑟−𝑑
2) Total annual inventory cost= √2𝐶0 𝐷𝐶𝑐 ( )
𝑟
𝑄𝑜𝑝𝑡
3) Optimum length of production run= 𝑑
𝑄𝑜𝑝𝑡
4) Production cycle time= 𝑟
Problem.
finds that when he starts production run he can produce 25000 bearings per day. The cost
of holding a bearing in stock for ayear is Rs. 2 and setup cost production run is Rs. 180.
Find the economic batch quantity and frequency of production run and production cycle
2𝐶0 𝐷 𝑟
EBQ= 𝑄𝑜𝑝𝑡 = √ (𝑟−𝑑)
𝐶𝑐
Qopt=30,000
𝑄𝑜𝑝𝑡
Frequency of production run= 𝑑
30000
=10000 = 3 𝑑𝑎𝑦𝑠
𝑄𝑜𝑝𝑡
Production cycle time= 𝑟
=3000000/25000
=1.2 days
UNIT-5
WAITING LINES
INTRODUCTION
A Flow of customers from infinite/finite population towards the service facility form a queue
(waiting line) on account of lack of capability to serve them at a time. The queue may be of
person waiting at a doctor’s clinic or at railway booking office, we wait to eat at restaurant and
this waiting phenomenon is not limited to human beings only. Jobs wait to process on machine,
cars and other vehicles wait at traffic signals. In the absence of a perfect balance between the
service facilities or for the customers, waiting is required either of the services facilities or for
the customer arrival.
By the term customer we mean the arriving unit that requires some service to be performed.
The customers may be persons, machines, vehicles etc. Queue stands for number of customers
waiting to be served. The queue does not include the customer being served. The process or
system that performs the service to the customers is termed by service channel or service facility.
The subject of Queuing is not directly concerned with optimization (max/min), rather if
attempts to explore understand and compare various queuing situations and thus indirectly
achieves optimization approximately.
2. Queue Discipline
3. Service Mechanism
1. Input process: - This element of Queuing system is concerned with the pattern in which the
customers arrive for service. Input source can be described by following three factors.
a) Size of the Queue :- If the total number of potential customers requiring service are only few,
then size of the input source is said to finite .If the potential customers requiring service area
sufficiently large in number, then the input service is considered to be infinite. When more than
one arrival is allowed to enter the system simultaneously the input is said to be occur in bulk or
batches.
b) Arrival Distribution: - If the pattern at which customers arrive at the service system or time
between successive arrivals ( inter arrival times ) is uncertain , the arrival pattern is measured by
either mean arrival rate or inter arrival time. These characterized by the probability distribution
associated with this random process. The arrival rate follows Poisson distribution and inter
arrival time follow an exponential distribution.
c) Customers behavior: - It is also necessary to known the reaction of customer upon entering
the system. A system be decide to wait no matter how long the queue becomes or if the queue is
too long to suit him may decide not to enter it .If a customer decides not to enter the queue
because of its huge length ,he is said to have balked. A customer may enter the queue but after
some time loses patience and decides to leave. In this case he is said to have reneged. In the case
when there are two or more queues, customer may move one queue to another for his personal
economic gains that is jockey for position.
2. Queue Discipline: - This is referred to the order in which the numbers of the queue are
selected for service. It may be first come first served, last come first served, Service in random
order or priority service.
First come First server (FCFS): - The order of service of customer is in the order of their arrival
or FIFO.
Last come first served (LCFS):- The order of service is the reverse of the order in which the
customers arrived (LIFO).
Ex:- Loading and unloading a truck or godown,wearing socks and shoes, dressing a shirt and
coat over it.
Served in random order (SIRO):- The selection of customers for service is done a random way.
Ex: - Priority given at APSRTC reservation counter who buy a CAT Card.
3. 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 services. If there is infinite number of servers then all the customers are served
instantaneously on arrival and there will be no queue.
If the number of servers is finite then customers are served according to a specific order.
a)Single queue one serve:- One queue one service channel, where in the customers waits till the
service point is ready to take him in for servicing.
b) Single queue several servers: - The customers wait in a single queue until one of the service
channels is ready to take them in for servicing.
c) Several queues one server: - There are several queues and the customer may join any one of
these but there is only one service channel.
d) Several servers: - When there are several service channels available to provide service much
depends upon their arrangements. They may arrange in parallel or in series or a more complex
combination of both, depending on the design of the systems service mechanism.
For series channels, it consists of multiple servers which are connected in series.
4) Capacity of the system: - The source from which customers are generated may be finite or
infinite. A finite source limits the customers arriving service. i.e., there is a finite limit to
maximum queue size. The queue can also be viewed as one with force balking where a customer
is forced to bulk if he arrives at a time when queue size is at it limit. Alternatively an infinite
source is forever abundant as in the case of telephone calls arriving at a telephone exchange.
Some of the operational characteristics of queuing system that are of a general interest for
the evaluation of the performance of an existing queuing system and to design a new system are
as follows.
1. Expected number of customers in the system denoted by E (n) or 𝐿𝑠 is the average number of
customers in the system, both waiting and in service. Here n= number of customers in the
queuing system.
2. Expected number of customers in the queue denoted by E (m) or 𝐿𝑞 is the average number of
customers waiting in the queue. Here m=n-1 i.e., excluding the customer being served.
3. Expected waiting time in the system denoted by E (v) or 𝑊𝑠 is the average number of
customers in the system .It is generally taken to be waiting time plus servicing time.
4. Expected waiting time in the queue denoted by E (w) or 𝑊𝑞 is the average time spent by
customers in the queue before the commencement of his service.
ƛ
5. The server utilization factor (or busy period) denoted by 𝜌(= 𝜇 ) is the proportion of time that
a server actually spends with the customers. Here ƛ stands for the average number of customers
arriving per unit of time and 𝜇 stands for the average number of customers completing service
per unit of time.
Terminology
1. Customers: - The persons or objects that required certain service are called customers.
2. Server: - The person or an object or a machine that provides certain defined service is known
as server.
3. Service: - The activity between server and customer is called service that consumes some
times.
5. Arrival: - The process of customer coming towards service facility or server to receive certain
service is arrival.
6. Mean rate of Arrival: - The average number of customers arriving per unit time is called
mean rate of arrival and it is denoted byƛ.
8. Rate of service: - It is the average number of customers served per unit time and is denoted
by𝜇.
The representation of Queuing model as given below. It is introduced by D.G. Kendall and
then by A. Lee.
(a / b / c): ( d / e / f )
d: service discipline
Symbols for d
Symbols for e
Let ƛ = mean or expeted number of arrivals per time period (mean arrrival rate)
4. Probability that there are more than or equal to k units in the system i.e., queue ≥ k
ƛ ƛ
𝑃(𝑛 ≥ 𝑘) = (𝜇)𝑘 Also 𝑃(𝑛 ≥ 𝑘) = (𝜇)𝑘+1
ƛ 𝑘 ƛ
𝑃(𝑛 = 𝑘) = (𝜇) (1 − µ)
ƛ
5. Expected or average number of units in the system or system length 𝐿𝑠 = 𝜇−ƛ
6. Expected or average number of units in the Queue waiting for service or length of queue
ƛ2
𝐿𝑞 = µ(µ−ƛ)
µ
7. Expected or average length of non-empty queue 𝐿𝑞 1 = µ−ƛ
1
8. Expected or average waiting time in the system 𝑊𝑠 = 𝜇−ƛ
ƛ 1
9. Expected or average waiting time in the queue 𝑊𝑞 = µ (𝜇−ƛ)
Problems
1. Arrival of cars at a filling station is considered to be a Poisson variant with an average time of
4 minutes between one arrival and the next. The length of filling petrol is assumed to be
distributed exponentially with mean of 0.05 hours. Find
V) The new arrival rate if the management of petrol pump is determined to install second pump
is determined to install second pump under the condition that the average waiting time with the
present exceeds 12 minutes.
60
Sol: Arrival Rate ƛ = = 15 𝑐𝑎𝑟𝑠/ℎ𝑜𝑢𝑟
4
60
Service Rate μ = = 20 𝑐𝑎𝑟𝑠/ℎ𝑜𝑢𝑟
3
Hence the new arrival rate is same as the given arrival rate.
Method-II:
µ: service rate
𝑀! ƛ 𝑖
1. Probability that the system shall be idle 𝑃0 = [(𝑀−𝑖)! (𝜇) ]−1
ƛ 𝑛 𝑀!
2. Probability that there n customers in the system 𝑃𝑛 = 𝑃0 (𝜇) (𝑀−𝑛)!
if 0<𝑛<𝑚
𝑃𝑛 = 0 If 𝑛>𝑚
ƛ+𝜇
3. Expected length of the queue 𝐿𝑞 = 𝑀 − (1 − 𝑃0 )
ƛ
𝜇
4. Expected number of customers in the system 𝐿𝑠 = 𝑀 − ƛ (1 − 𝑃0 )
1 𝑀 ƛ+𝜇
5. Expected waiting time of a customer in the queue 𝑊𝑞 = 𝜇 [1−𝑃 − ]
0 ƛ
1 𝑀 ƛ+𝜇 1
6. Expected time a customer spends in the system 𝑊𝑠 = 𝜇 [1−𝑃 − + 1] = 𝑊𝑞 + 𝜇
0 ƛ
Problem
1. In a factory having 5 machines, breakdown occurs randomly and the average time
between the breakdowns is 2 days. Assuming that the repairing capacity of the workman
is one machine a day and the repairing is distributed exponentially. Find
i) Probability that the service facility is idle.
ii) Probability of various numbers of machines to be and being repaired.
iii) Expected length of the queue.
iv) Expected number of machines waiting to be and being repaired.
v) Expected time a machine shall wait in the queue to be repaired.
vi) Expected time a machine shall be idle for reason, for waiting to be and being repaired.
Sol: M = 5 machines
ƛ
∴ 𝜌 = 𝜇 = 0.5
Since 𝑃0 is basic to all formulae for this model, we shall first determine its value.
i 𝑀! ƛ 𝑖 𝑀! ƛ 𝑖
(𝑀 − 𝑖)! ( ) ( )
𝜇 (𝑀 − 𝑖)! 𝜇
0 1 1.0000 1
1 5 0.5000 2.5
2 20 0.25 5.00
3 60 0.125 7.5
4 120 0.0625 7.5
5 120 0.03125 3.75
27.25
𝑀! ƛ 𝑖
𝑃0 = [ ( ) ]−1 = 0.0367
(𝑀 − 𝑖)! 𝜇
In this system the maximum size of the queue permissible is N. This model is also known
as finite queue model.
2. The service time has exponential distribution with an average service rate.
7. The waiting space available for the customer in the rate queue is finite given by N.
(1−𝜌)𝜌𝑛
2. Probability that there are n units in the system 𝑃𝑁 = = 𝑃0 𝜌𝑛 , 𝜌 ≠ 1, 𝑛 = 0,1,2, − −
1−𝜌𝑁+1
−−−−𝑁
𝜌 (𝑁+1)𝜌𝑁+1
3. Expected or average number of customers in the system 𝐿𝑠 = 1−𝜌 − 1−𝜌𝑁+1
𝐿
4. Expected or average waiting time in the system 𝑊𝑠 = ƛ1𝑠 where ƛ1 = ƛ(1 − 𝑃𝑁 ) effective
arrival rate
1
5. Expected or average waiting time in the queue 𝑊𝑞 = 𝑊𝑠 − 𝜇
1
Where 𝜌 = 1 → 𝑃𝑛 = 𝑁+1
𝑁−1 𝑁(𝑁−1)
𝐿𝑠 = , 𝐿𝑞 =
2𝜇 2(𝑁+1)
Problem
1.In a bank with single server , there are 3 chairs for waiting customers. on a average one
customer arrive every 7.5min and each customer takes 6 min for getting served. Making suitable
assumptions. Find
60
𝑆𝑒𝑟𝑣𝑖𝑐𝑒 𝑅𝑎𝑡𝑒 𝜇 = = 10 𝑝𝑒𝑟 ℎ𝑜𝑢𝑟
6
𝑃𝑁 = 𝑃0 𝜌3 = 0.174
𝜌 (𝑁+1)𝜌𝑁+1
𝐿𝑠 = 1−𝜌 − = 1.225
1−𝜌𝑁+1
𝑠 𝐿
𝑊𝑠 = ƛ(1−𝑃 = 0.185
𝑁)
1
𝑊𝑞 = 𝑊𝑠 − 𝜇 = 0.085ℎ𝑟𝑠 = 5.1𝑚𝑖𝑛
Method-IV:
Multi channel queuing theory treats the conditions in which there are several service
stations in parallel and each element in the waiting line can be served by more than one station.
Each service facility is prepared to deliver the same type of service. The new arrival selects one
station without any external pressure.
When a waiting line is formed, a single line usually breaks down into shorter line in front of
each service station. The arrival rate ƛ and service rate 𝜇 are mean values from Poisson and
exponential distribution respectively.
Service discipline is first come first served and customers are taken from a single queue. If
there is any empty channel it is filled by the next customers in line.
When 𝑛 < 𝑘 there is no queue and all arrivals are being served and rate of service will be n.µ
as only n channels are busy. When n=k all channels will be working.
When 𝑛 > 𝑘 there will be (n-k) persons in the queue and rate of service will be kµ as all the k
channels are busy.
ƛ 𝑘
ƛµ( )
µ
𝐿𝑞 = (𝑘−1)!(𝑘𝜇−ƛ)2 𝑃0
ƛ 𝑘
ƛµ( ) ƛ ƛ
µ
𝐿𝑠 = (𝑘−1)!(𝑘𝜇−ƛ)2 𝑃0 + µ = 𝐿𝑞 + µ
6. Average waiting time of an arrival (expected time a customer spends in the queue waiting for
service)
ƛ 𝑘
µ( ) 𝐿𝑞
µ
𝑊𝑞 = (𝑘−1)!(𝑘𝜇−ƛ)2 𝑃0 = ƛ
ƛ 𝑘
ƛµ( ) 1 𝐿𝑠
µ
𝑊𝑠 = (𝑘−1)!(𝑘𝜇−ƛ)2 𝑃0 + 𝜇 = ƛ
ƛ
8. Utilization factor 𝜌 = 𝑘𝜇
Problem
1. A tax consultant firm has 4 service stations (counters) in its office to receive people who have
problems and complaints about their income, wealth and sales taxes. Arrivals average
100persons in 10 hours service day. Each tax adviser spends an irregular amount of time
servicing the arrivals which has been found to have an exponential distribution. The average
time is 20min. calculate 𝑃0 , 𝐿𝑞 , 𝐿𝑠 , 𝑊𝑞 , 𝑊𝑠 , 𝑃(𝑛 > 𝑘).
100
Sol: G.T ƛ = = 10𝑝𝑒𝑟 ℎ𝑜𝑢𝑟
10
60
𝜇 = 20 = 3𝑝𝑒𝑟 ℎ𝑜𝑢𝑟
k=4
ƛ 10
𝜌 = 𝑘𝜇 = 12
𝑘−1
1 ƛ 𝑛 1 ƛ 𝑘 1
𝑃0 = [∑ ( ) + ( ) ( )]−1 = 0.0208
𝑛! μ 𝑘! μ ƛ
𝑛=0 1−
𝑘μ
ƛ 𝑘
ƛµ( )
µ
𝐿𝑞 = (𝑘−1)!(𝑘𝜇−ƛ)2 𝑃0 = 3.24 𝑐𝑢𝑠𝑡𝑜𝑚𝑒𝑟𝑠
1 ƛ 1
vi) Probability that customers have to wait 𝑃(𝑛 ≥ 𝑘) = 𝑘! (µ)𝑘 ƛ 𝑃0 = 0.618
[1− ]
𝑘µ
Dynamic Programming
Identify the decision variables
↓
Specify the objective function
↓
Decompose in to smaller problems (Stages)
↓
Identify the static variables
↓
State the general recursive relations. Also decide forward /forward
↓
Calculate required values of return function by appropriate formulae at each
stage
↓
Repeat and determine the overall optimal policy
Introduction
In Dynamic programming a large problem is split into smaller sub problems each of which
involve in a few variables. These sub problems are sequentially optimized so as to make the total
problem optimal.
Discrete and continuous, deterministic as well as probabilistic models can be solved by this
method. Thus Dynamic Programming method is very useful in obtaining optimal solution of
various problems such as inventory, replacement, allocation, linear programming, Reliability
calculations etc. A Single constraint problem is relatively simple, but in the problem of more
than two constraints more complexities may appear.
2. Stage; - A large problem is decomposed in to smaller and sequential sub problems. Each
sub problem is referred to as a stage where a decision is called for. Thus each stage can be
considered to have a starting and ending. The ending of one stage will be the beginning of its
next stage.
3. State:-Any stage just before starting or just after completion will have certain status called
the state and the variable that links up to stages is called state variable.
4. Return function: - The decision of a stage described in the form of an algebraic equation
with state variables to explain the worth or benefit or cost is known as Return function or
transformation equation.
The basic characteristics of the Dynamic Programming Problem are given below.
1. The problem can be sub divided into sub problems referred to as stages with policy
decision at each stage.
2. A stage is a device to sequence the decisions. This means that the sequential summation
of these decisions obtained from optimal solution of sub problems will be the part of the
optimal solution of the total problem.
3. Every stage consists a number of states associated with it. The states are different
possible conditions in which system may find itself at that stage of the problem.
4. Decision at each stage convents the current stage into state associated with the next stage.
5. The state of the system at stage is described by a set of variables called state variables.
6. When the current state is known, an optimal policy for the remaining stages is
independent of the policy of the previous stages.
7. To identify the optimum policy for each state of the system. A recursive equation is
formulated with n stages remaining given the optimal policy for each state with (n-1).
8. Using recursive equation approach each time the solution moves either backward or
forward by stage obtaining the optimum policy of each state for that particular stage, till
it attains the optimum policy beginning at final or initial stage respectively.
Problems:
1. A Medical representative located at city 1 has to travel to city 10. He knows the
distance of alternative routes from city 1 to city 10 and has drawn the network
map based on the distance between the cities as in the following table. Draw the
network and find the shortest possible route. Also, find the shortest routes from
any city to city 10. (March-2017 R-13)
From To Corresponding
city city distance in km
1 2,3,4 4,6,3
2 5,6,7 7,10,5
3 5,6,7 3,8,4
4 5,6,7 6,10,5
5 8,9 4,8
6 8,9 3,8
7 8,9 8,4
8 10 7
9 10 9
The given problem can be divided into 5 stages as shown in the figure. It is sived in the
tabular form method as shown below.
At stage 1 if the medical representative is at city 10, it is the destination itself, he need
not travel any more.
Stage 2: At stage 2 i.e if he is at city 8 or city 9 the shortest path calculation is shown
below.
Decisions Destinations Minimum Optimal
State Stage distance Cumulative distance decision
variable(S1) distance
Entering 8 7 7 7 8-10
state 9 9 9 9 8-10
Stage-3:
Stage-5: With the optimal distance decisions we now proceed to calculate decision table of
stage-5
Distance(km): 6+3+4+7=20 km
Solution set-2: Sequence(City) 1-4-5-8-10
Distance: 3+6+4+7=20 km
We can thus state that medical representative may have to travel minimum number of 20 km
from city 1 to city 10.
2) Consider the problem of allocating salesmen to three zones in such a way that the total
profits of the company are maximized. The following table gives the details of the zones,
number of salesmen and the corresponding profits.
Sol. Stage-1:
x 0 1 2 3 4 5 6 7 8 9
F1(x) 30 45 60 70 79 90 98 105 100 90
This table gives the profit corresponding to different salesmen allocated to zone 1
F1(x)= f1(x)
Stage-2(Zones1 &2):
S= f2 (x)+ F1(A-x)
This equation can be applied to determine the optimal distribution of any number of salesmen
(A=9,8,7,6,5) as shown in table
Zone 1 0 1 2 3 4 5 6 7 8 9 No. Of
sales men
Zone 2 30 45 60 70 79 90 98 105 100 90 Profits
0 35 65 80 95 105 114 125 133 140 135 125
9 100 130
The maximum profit for each value of A are represented. The above table shows the profits for
all combinations.
A 0 1 2 3 4 5 6 7 8 9
S 65 80 95 105 115 125 135 143 154 163
0+0 0+1 0+2 1+2 1+3 0+5 1+5 3+4 3+5 6+3
(or) (or)
0+3 1+6
We have
9 110 175
A= (x1+x2)+x3
A 0 1 2 3 4 5 6 7 8 9
S 107 122 137 149 159 169 179 190 200 210
0+0 0+1 0+2 1+2 1+3 1+4 1+5 5+2 5+3 5+4
Thus for A=9 we find that the profit is maximum for x1+x2=4 and x3=5. From stage II we find
that for A=4, the profit has maximum value 115 with x1=3, x2=1.
Therefore the optimal distribution of salesmen is
Zone-1→3.
Zone-2→ 1
Zone-3→ 5
3.A truck can carry a total of 10 tons of a product. Three types of product are available for
transport. Their weight and values are given below. Determine the loading such that the total
value of the cargo is maximum (At least one unit of each type must be loaded)
Ans. We have to determine how many units of A,B,C is to be loaded. We take the loading of
type C as stage-1.
Stage-I:
State x1
No. of units Optimal value
(Weight)
1 2 3 Value Decision
C11=60 C12=120 C13=180 F1A)
W=2 W=4 W=6
2 60 - - 60 1
3 60 - - 60 1
4 60 120 - 120 2
5 60 120 - 120 2
6 60 120 180 180 3
7 60 120 180 180 3
Stage-II:
State x2
No. of units Optimal value
(Weight)
1 2 3 Value Decision
C21=50 C22=100 C33=150 F2(A)
W=2 W=4 W=6
4 50+6 100 - 110 1
5 50+60 100 - 110 1
6 50+120 100+60 150 170 1
7 50+120 100+60 150 170 1
8 50+180 100+120 150+60 230 1
9 50+180 100+120 150+60 230 1
F3(A)=Max{f3(x)+F2(A-x)}