Professional Documents
Culture Documents
UNIVERSITY OF KERALA
MGT 202
OPERATIONS RESEARCH
1
Syllabus for operation Research MBA
Unit II : Transportation models. North west control method (NWC), Least cost
method (LCM), Vogels approximation method, Hungarian Method, Minimisation of
cost
Unit III : Network models, critical path method (CPM), Programme evaluation
review technique (PERT) Crashing resource smoothing – Resource leveling
Unit IV : Queuing Theory M/M/I (System with single server) M/M/2 (system with
two servers). Monte carlosimulatation, Application in Queing + Inventory
Unit V : Decision theory : Decision making under risk and uncertainty,Game theory,
Decision tree ,ORsoftwares
References
2
Module 1
Operations research
Operations Research is the systematic and method oriented study of the basic structure,
functions and relationships in an organization.
Operation research is the application of scientific methods, techniques and tools to
problems involving the operation of a system so as to provide those in control of the
system with optimum solution to the problem
Operations research is scientific methods of providing executive department with a
quantitative basis for decision regarding the operations under the control.
Operation Research is an applied decision theory which involves a team approach of
experts from various disciplinessuch as scientists,Engineers ,Business executives
,psychologistsetc
Features (characteristics) of OR
1. System Orientation – All possible interactions in activities of organization
2. Interdisciplinary approach – Team of scientific from different faculties
3. Scientific approach :Uses scientific methods to solve complex problem.
4. Decision making : Management to take better decision
5. Optimization of objective: OR attempts to fund the basis and optimal solution to
problem using OR techniques. It tries to optimize a well defined function subject to
given constraints.
6. Mathematical Models and quantitative solution
OR uses models built by quantitative measurements of variables concerning a given
problem and drives a quantitative solution from the model.
7. Use of computer : OR often require a computer to solve the complex mathematical
model
3
Phases of operation Research
1. Formulating the problem: identifying defining a specifying the components of
model
2. Constructing the model : Objective function & constraints
3. Deriving the solution – Solution to the model
Analytical method, numerical method and simulation methods are used
4. Testing the validity : A model is said to be valid, if it can gave a reliable prediction
of system performance.
5. Controlling the solution : Control over solution of proper feedback of information
variables which deviated significantly
6. Implementing the results – Examination of solution to be adopted
OR techniques
1. Allocation model: Allocation of resources such that effectiveness is optimized.
Allocation problem can be solved by linear &non linearprogramming Technique
.Assignment models and Transportation models are special cases.
2. Sequencing : Placing items in a certain sequence
3. Waiting or queuing theory : Models involving waiting for services
4. Inventory models : These are models with regard to holding or storing resources aim
at optimum inventory level
4
5. Competitive strategy models (Game theory) : When two or more people competing
for a certain resources,. Game model is used to determine optimum strategy in a
competitive situation
6. Decision Theory: Concern with making sound decisions under conditions of
certainty, risks &uncertainty.
7. Network analysis :Involve determination of an optimum sequence of performing
certain operations concerning some job in order to minimize overall time or cost. Eg.
PERT, CPM, Gantt Charts
8. Simulation : Technique of testing a model which represent real life situation
9. Search models : The models concern itself with search problem
10. Replacement theory : These are models concerned with the situation that arise when
some items need replacement because the same may deteoriatewith time or break
down completely.
OR has a very great impact on Economics, management Engineering and other social and
behavioral service
Marketing function
5
2. Queing theory for determining the size of maintenance crews
3. Scheduling techniques for man power scheduling
5. Finance function
1. Integra programming for capital for budgeting
2. Portfolio analysis
3. Investment analysis
6. Purchasing
1. Material transfer
2. Optimal buying
3. Optimal reordering
7. Research and development
1. Control of R&D projects
2. Product introduction planning
Models of OR
Advantages of a model
6
5. It enables the use of high powered mathematical techniques to analyse the
problem
6. It helps in finding avenues for new research and improvement in the system
Disadvantages of model
Types of models
Descriptive Normative or
models optimization
model
Deterministic Probabilistic or
model stochastic
models
7
decision alternatives. In this model, it is possible to get information as to how one or more
factors change as a result of changes in other factors.
Models by structure
Iconic or physical models : They are pictorial representation of real system and have the
appearance of the real thing. Iconic models are obtained by enlarging or reducing the size of
the system. In other words they are images.
Eg .photographs, drawing model air planes, globes, map etc.
These models are easy to observe and difficult to manipulation
a. . Symbolic models
These model employ a set of mathematical symbol to represent the components of
real system. This symbolic models are some kind of mathematical equation or
inequalities reflecting the structure of system they represent.Eg : Inventory models,
Queuing models, allocation models ,sequencing models etc.
These models usually yield more accurate results under manipulation
b. Analogue models : In analogue models, one set of properties is used to represent
another set of properties. After the problem is solved, the solution is reinterpreted iin
terms of original system. Analogue models are easy to manipulate than iconic models
Eg.Control linear on a map are analogues of elevation as they represent the rise and
fall of heights
Deterministic models
In deterministic models, everything is defined and the results are certain
Eg. EOQ methods, economic lot size
Probabilistic model (Stochastic models)
In this model there is risk and uncertainty and thereforeinput and output variables
assume probability distribution. In this model at least one parameter or decision
variable is a random variable.
Eg. Game theory, if the strategy is not pure, we apply probability model
Static Model: In this model, time factor is not considered.
Eg. Inventory problem of determining EOQ
Heuristic Model : In this model, are used to explore alternative strategies which have
been over looked previously.
8
Linear programming (LPP)
Linear programming is one of the important operation research techniques used in business
and industry.
Linear programming is a mathematical technique which involves allocation of limited
resources in an optimal manner, on the basis of a given criterion of optimality
Linear programming deals with the optimization of a function of variables known as
objective function to a set of linear equalities/ inequalities. Optimization may be maximize or
minimize.
2. Additivity: Sum of resources used by different activities must be equal to the total
quantity of resources used by eachactivity for all the resources individually or
collectively
9
3. Divisibility : The solution need not be whole number
4. Certainity : Coefficients in the objective function and constraints are completely known
and do not change during the period under study.
Formulation of LPP
10
16 hours of time per day available on machine A and 30 hours on machine B. Profit gained
by the manufactures from a chair isRs. 1 and from a table is Rs. 5 respectively. Formulate
the problem into a LPP in maximize the total profit.
Solution
Let x1 be the number of chairs and x2 be the no of tables produced.
Profit from chair = 1 x x1
Profit from table = 5 x x2
Total profit Z = x1 + 5x2
Constraints
Machine B
Time required for chairs = 2 x x1 = 2x1
Time required for tables = 5 x x2 = 5x2
Total time required = 2x1+ 5x2
Available time = 16 hours
2x1 + 5x2 ≤ 16
Machine B
Time required for chairs = 6 x x1 = 6x1
Time required for tables = 0 x x2 = 0
Total time required = 6x1 + 0
Available time = 30 hours
6x1 + 0 ≤ 30
LPP :Maximize z = x1 + 5x2
Subject to 2x1 + 5x2 ≤ 16
6x1 ≤ 30
x1>0 x2 ≥0
2. A manufacturer of medicines is preparing a production plan on medicines A & B. There
are sufficient ingredients available to make 20,000 bottles of A and 40,000 bottles of B. But
there are only 45000 bottles into which either of the medicines can be put. Further more, it
takes 3 hours to prepare enough material to fill 1000 bottles of A, it takes one hour to
prepare enough material to fill 1000 bottles of B and there are 66 hours available for the
operation. The profit is Rs. 8 per bottles for A and Rs. 17 per bottle for B.
Formulate the problem as LPP
Solution
Let x1 be the no of bottles of medicine A to be produced
Let x2 be the no of bottles of medicine B to be produced
Objective function
Maximize Z = 8x1 + 17x2
Constraints ( 3 types)
1. Availability of bottles
2. Availability of hours
3. Availability of ingredients
First constraints
11
X1 + x2 ≤ 45000 --------(1)
Second constraint
+ ≤ 66
Solution
Decision variable
Let x1 and x2 be the number of units of liquids and dry products
Objective function
Minimise Z = 3x1 + 2x2
Constraints for 3 chemicals
5x1 + x2 ≥ 10
2x1 + 2x2 ≥ 12
X1 + 4x2 ≥ 12
X1 x2 ≥0
Problem 4
A company manufactures two products A and B. These products are processed in the same
machine. It takes 10 minutes to process one unit of product A and 2 minutes for each unit of
product B and the machine operates for maximum 35 hours in a week. Product A requires 1
kg and 0.5 kg of raw material per unit the supply of which is 600 kg per week. Market
constraint on product B is known to be 800 unit every week. Product A costs Rs. 5 per unit
sold at Rs. 10. Product B costs Rs. 6 per unit and can be sold in the market at a unit price of
Rs. 8.
12
Determine the number of units of A and B per week to maximise the profit
Solution
Objective function
Max Z = 5 x1 + 2x2
LPP
X1 + 0.5 x2 ≤ 600
X2 >800 x1 x2 ≥ 0
Diet problem
Problem 5
Objective of diet problem is to ascertain the quantities of certain foods that meet the
nutritional requirement at minimum cost. The food articles considered are milk, beef
and eggs and the nutritional requirements are vitamins A, B & C. The number of
milligrams of each of these vitamins contained within a unit of each food is given
below.
Solution.
Minimum daily
Vitamin Gallon of milk Pound of beef Dozen of egg
required
A 1 1 10 I mg
B 100 10 10 50 mg
C 10 100 10 50mg
Cost Rs 10 1.10 0.50
13
Formulate a LPP for the problem
Solution
Let x1 be the gallons of milk x2 be the pound of beef and x3 be the dozen of egg to be
included in the daily diet
Minimise
Subject to
x1 + x2 + 10x3 ≥ 1 (Vitamin A)
x1 x2 x3 ≥ 0
Activity Questions
6 A company manufactures two types of products P1 & P2. Each product uses lathe and
milling machine. The processing time per unit of P1 on the lathe is 3 hours and on the
milling machine is 4 hours. The processing time per unit of P2 on the lathe is 10 hours and
on the milling machine is 4 hours. The maximum number of hours available per week on
the lathe and milling machine are 60 hours and 40 hours respectively. Also the profit per
unit of selling p1 and p2 are Rs6 and Rs8 respectively.
Formulate a linear programming model to determine the production volume of each of the
products such that total profit is maximized.
Weaving machine
Process number Processing cost / sqm ® Max daily capacity (sq)
Set up cost (Rs)
1 150 15 2000
2 240 10 3000
3 300 8 3500
14
The daily demand for cost for its white polyester fabric is 4000 sqm. The company
production manager wants to make a decision concerning which combination of
production processes is to be utilized to meet the daily demand for cost and what
production level of each selected production process to be operated to minimize the total
product costs
A company produces two types of cow boy hats. Each hat of the first type requires twice
as much alabour time as the second type. If all the hats are of the second type only, the
company can produces a total of 500 hats a day. The market limits daily sales of first type
and second types to 150&200 hats a day. Assuming that profit per hat areRs. 8 for type
I and Rs. 5 for type 2. Formulate the problem as a LP model in order to maximize profit.
15
Solution
1. A objective function
2. Set of constraints
3. Non negative restrictions
In a graphical method, the problem is solved by the method of graph drawn in a two
dimensional plane. Since only two variables can be represented in a two dimensional
plane, the graphical solution can be used only when the objective function has just two
variables x1 and x2
When the objective function has three or more variables, graphical method cannot be
used to solve the problem
Solution is the coordinate of vertex which optimizes the objective function and
corresponding values of the objective function
16
X1 x2 ≥0
Solution
2x1 + 3x2 = 1500 Put x1 = 0 , x2 = 500
Put x2 = 0, x2 = 750
Points are (0,500)(750,0)
3x1 + 2x2 = 1500 Put x1 = 0 , x2 = 750
Put x2 = 0, x2 = 500
Points are (0,750), (500,0)
3x1 + 2x2 = 1500 Put x1 = 0 , x2 = 750
Put x2 = 0, x2 = 500
3x1+ 2x2 = 1500 Put x1 = 0 , x2 = 750
Put x2 = 0, x2 = 500
Points are (0,750) (500,0)
Feasible region lies to the left of three quadrant lines. The feasible region is marked as the
shaded region.Optimal
region solution lies at the corner points of the boundary of the feasible
solution
1. M1 (0, 400) z = 50x1 + 60x2
Z = (50 x 0) + 60 x 400 = 24000
2. M2 (150,400) Z = 50 x 1580 + 60 x 400 = 31500
3. M3 (300, 300) Z = 50 x 300 + 60 x 300 = 33000
4. M4 (400, 150) Z = 50 x 400 + 60 x 0 = 20000
17
8x1 + 4x2 ≥80
X1, x2 ≥ 0
Solution
5x1 +4x2 = 200put
200 x1=0 x2= 50
Put X2 = 0 x1= 40
Points are (0,50) (40,0)
3x1 + 5x2 = 150 Put x1 =0 x2 = 30
Put X2=0, x1 = 50
Points are (0, 30) (50,0)
5x1 + 4x2 = 100 put x1 = 0, x2 = 25
Put X2 = 0 x1 = 20
Points are (0, 25),
25 (20, 0)
8x1 + 4x2 = 80 put x1 = 0 x2 = 20
Put X2 = 0, x1 = 10
Points are (0,20) (10,0)
= 138.4
18
X – 2y ≤ 200
2x + y ≥ 100
Y ≥ 200 x, y > 0
Solution : The equation can be converted to inequalities
X + y = 300
X – 2y = 200
X + 2y = 200
Y = 200
First constraint x + y = 300
Put x =0, y= 300
Put y=0, x= 300
Points are (0, 300) (300, 0)
Second constraint x – 2y = 200
Put x = 0, y= -100
Put y = 0, x= 200
Points are (0,-100)(200,0)
Third constraint
2x+y=100
Put x=0 y=100
Put y=0, x=50
The points are (0,100)(50,0)
19
The critical points are M1(0,300)
2qZ= x-3y=
3y= 0-300x3= -900
M2 (0, 100)
M3 = (50,0)
Z = 50 – 3 x 0 = 50
M4 = (75,- 75)
Z = x- 3y 75 – 3 x- 75 = 300
M5 = (200, 0)
Z = x- 3y = 200-
200 3x 0 = 200
M6 = (265, 35)
Z = 265-33 x 35 = 265 – 105 = 160
Maximum value of z at the point M4 (75, 100)
Where z = 300
Maximum value of objective function = 300
Minimisation case problems
Problem 4
Solve graphically the LP problem
Minimize z = 1000 x1 + 800 x2
Subject to 6x1 + 2x2 ≥ 12
2x1 + 2x2 ≥ 8
4x1 + 12x2 ≥ 24
20
x1 ≥
≥0, x2 ≥ 0
Solution
If x1 = 0, x2 = 6
If x2 = 0, x1= 2
If x1 = 0, x2 = 4
If x2 = 0, x1 = 4
If x1 = 0, x2 = 2
If x2 = 0, x1 = 6
21
M1 (6,0) Z = 100x1 + 800 x2
= 100 x 6 + 0 = 6000
Problem 5
A Small scale manufacturer has production facilities for producing two different products.
Each of the product requires three different operations. Grinding , assembling and testing .
product I requires 15, 20 and 10 minutes to grind, assemble and test respectively where as
product 2 requires 7.5 , 40 and 45 minutes for grinding, assembly and testing. The production
run calls for at least 7.5 houses of grinding time, at least 20 hours of assembly time and at
least 15 hours of testing time. If product I costs Rs. 60 and product II costs Rs. 90 to
manufacture, determines the number of each product the firm should produceinorder to
minimize the cost of operation?
Minimize Z = 60x1 + 90 x2
Subject to 15x1 + 7.5x2 ≥ 7.5 x 60
20x1 + 40 x2 ≥ 20 x 60
10x1 + 45x2 ≥ 15 x 60
X1 > 0 x2 ≥ 0
Converting all the constraints as equations
15x1 + 7.5 x2 = 450
20x1 + 40 x2 = 1200
10x1 + 45x2 = 900
X1 = 0, x2 = 0
Ist constraint 15x1 + 7.5x2 = 450
X1= 0, x2 = 60
X2 = 0, x1= 30
Points are (0,60),(30,0)
2nd constraint
20x1 + 40x2 = 1200
Put X1 =0, x2 = 30
22
Put X2 = 0, x1= 60
Points are (0, 30) (60,0)
(60,
rd
3 constraint
10x1+ 45 x2 = 900
Put X1 = 0, x2 = 20
Put X2=0, X1=90
Points are (0,20),(90,0)
Activity Questions
23
Problem6 Solve the LPP graphically
Max Z = 5x + 8 y
Subject to 3x + 2y ≤ 36
X+ 2y ≤ 20
3x + 4y ≤ 42
X,Y ≥0(AnsMaxZ=82 when X=2,Y=9)
Problem 7
Egg contains 6 units of vitamin A per gm and 7 units of vitamin B per gm and the cost 12
paise per gm. Milk contains 8 units of vitamin A per gm and 12 units of vitamin B per gm
and costs 20 paise per gm. The daily minimum requirement of Vitamin A and vitamin B are
100 units and 120 units respectively. Find the optimal product mix.
Solution
Min Z = 205 x1 = 15, x2 = 1.25
Problem 8
Minimize Z= 25x1 + 22x2
Subject to constraints X1 + x2 ≥ 10
300 x1 + 400 x2 ≥ 3400
80x1 + 50x2 ≥ 680
X1 , x2 ≥ 0
Problem 9
Solve graphically
Maz Z = x1- 2x2
Subject to contraints– x1 + x2 ≤ 1
3x1 + 2x2 ≥ 12
0≤ x1 ≤ 5
2 ≤ x2 ≤ 4
Ans z = 1 where x1 = 5 x2 = 2
Problem 10
Old hens can be bought for Rs. 2 each but young ones cost Rs. 5 each. The old hens lay 3
eggs per week and the young ones 5 eggs per week each egg being worth 30 paise. A hen
costs Rs. 1 per week to feed., If I have only Rs. 80 spend for hens, how many of each
kind should I buy assuming that I cannot house more than 20 hens.
Some special cases of solutions
1.Infeasible : When no solution is obtained for a LPP satisfying all the constraints, there
exists no feasible solution that is if the feasible region of LPP is empty the solution is
infeasible
2. Unbounded : If in an LPP the solution of a variable can be made infinitely large
without violating the constraints, the solution is unbounded. That is feasible region is
unbounded for such a LPP
24
3. Alternative: If there are more than one optimum solution for the decision variable ( the
value of the objective function remaining the same) we say there are alternative or
multiple solution to LPP.
Problem 11
SDE UNI Feb 2016(5 Marks)
A firm engaged in producing two models A, B performs three operations. Painting,
assembling and testing. The relevant data are as follows (5 marks)
25
Required area is O A B C D
Required points are O, A, B, C, D
26
Problem 12
Required area is O A B C D
27
Point Value of objective function
Values
Values Z = 5x1+7x2
O (0,0) 0
A (0, 3) Z = 7*3 =21
B (1.5, 2.5) Z = 5*1.5+7 = *2.5 = 25
C (2.1, 1.8) Z = 5*2.1+7*1.8 = 23.1
D (3.5,0) Z = 5*3.5 = 18.5
Maximum value of objective function Z=25, When x1 = 1.5, x2 = 2.5 at point B
Simplex method
Simplex method is a linear programming technique in which we start with a certain
solution which is feasible. We improve the solution in a number of consecutive stages
until we arrive at an optimal solution.
For arriving at the solution LPP by this method, the constraints and objectives function
are presented in a table known as simplex table.
Main steps involved in simplex algorithm
1. Find an initial basic feasible solution of LPP
2. Test whether it is an optimal solution or not
3. If not optimal; improve the first trial basic feasible solution by a set of rules
4. Repeat steps 2 and 3 till an optimal solution is obtained
Feasible solution
A feasible solution to a LPP is the set of values of variables which satisfies all the
constraints and non negativerestrictions of the problem
Optimal solution
A feasible solution to a LPP in which the vector associated to non zero variables are
linearly dependent is known as basic feasible solution.
28
If all ∆js are positive solution is optimal.If any of the ∆j s are negative, continue from
steps 5 onwards.
5. Select the entering variables which has the highest negative value called key column
.A column which has highest negative value in a maximization problem and calculate
the key row which is the row correspond to minimum ratio column in the table. These
values are calculated by dividing the values in the Bj column by the corresponding
values in the key column
The values in the ratio column are known s exit ratios or limiting ratio. The variables
with the smallest non negative ratio is the exit variable
6. Calculate the new values for the new table value corresponds top values in the key
rows are obtained by dividing each value in the key row by the key number
7. New values can be calculated in remaining rows are calculated as follows.
New row = (old row values to be changed ) - (Corresponding new value for key row)
(value corresponding to values in row to be changed)
Repeat steps 4,5,6 till optimality is reached
I. Problem 1
X1 + x2 ≤2
5x1 + 2x2 ≤ 10
X1 + x2 + S1 = 2
5x1+ 2x2 + S2 = 10
Simplex table
29
Zj-Cj -5 -3 0 0 0
All Zj-Cj are negative the solution is not optimal
New table
2= =2 =1 =2 =2 =0
0/0 = 0 0/1 = 0
Pivottedoperation
New row = (old row value to be changed ) – (Corresponding new value for key rows)
– value corresponds to key value to be changed)
10 = 10- (2 x 5) = 0
5 = 5 – (1 x 5) = 0
2 = 2- ( 1 x 5) = -3
0= 0-(1 x 5) = -5
1 = 1- (0 x 5) = 0
0 = 0 – (0 x 5) = 0
New values
12 = 12- (2 x 3) = 6
3 = 3- (1x 3) = 0
8 = 8- (1 x 3) = 5
0 = 0- (1 x 3) = - 3
0 = 0- (0x 3) = 0
1= 1- (0x3) = 1
0
Value of Value of
5 3 0 0
Cj basic basic
x1 x2 S1 S2 S3
variable variable
5 X1 2 1 1 1 0 0
0 S2 0 0 -3 -5 1 0
0 S3 6 0 5 -3 0 1
30
Zj 10 5 5 5 0 0
Zj-Cj 0 2 5 0 0
0 S1 50 2 4 1 0 0 = 50
0 S2 100 2 5 0 1 0 = 20
0 S3 90 2 3 0 0 1 = 30
31
Zj 0 0 0 0 0 0
Zj-Cj -4 -10 0 0 0
Smallest non negative value corresponds to row S2. Hence x2 enters and S2 leaves.
Basic 4 10 0 0 0
variable X1 X2 S2 S2 S3
X2 =
20 0.4 1 0 0.2 0
10
X3 =0 30 0.8 2 0 0.6 1
Zj 4 10 0 2 0
Zj-Cj 0 0 0 2 0
1/5=0.20, 0/5=0
Pivottedoperations
32
Solve the following using simplex method
Subject to constraints
3x1 – x2 + 3x3 ≤ 7
-2x1+ 4x2 ≤ 12
X1, X2, X3 ≥0
Solution
Solution
Subject to constraints
3x1- x2 + x3 ≤ 7
-2x1+4x2 ≤ 12
Subject to constraints
3x1- x2 + x3 + s1 = 17
-2x1+ 4x2 + S2 = 12
Cj -1 3 -2 0 0 0
Value
Basic
basic X1 X2 X3 S1 S2 S3 Ratio
variable
variable
33
7/1
0 S1 7 3 -1 1 1 0 0
=7
12/4
0 S2 12 -2 4 0 0 1 0
=3
10/3
0 S3 10 -4 3 8 0 0 1
=3.33
Zj 0 0 0 0 0 0 0 3.3
Zj – Cj 1 -3 2 0 0 0
X2 entries S2 leaves
New table
= 0.25, =0
Pivoted operation
7 = 7-(-1x3) = 10 10 = 10-(3x3) = 1
3 = 3-(-1x )= -4 = -4-(- x 3) = 0
-1 = -1-(1x-1) =0 8 = 8- (0 x 3) = 8
1=1-(0x -1)=1 0=0-(0x3) = 0
1= 1-(0x-1)=1 0= 0 –( x 3) = - -
1=1-(0x-1) =1 1 = 1- (0 x3) = 1
0=0-( x -1) =
0=0-(0 x -1) = 0
Cj -1 3 -2 0 0 0
5 1
X1 X2 X3 S1 S2 S3 Ratio
2 4
0 S1 10 0 1 1 0 10/5/2=4
34
3
−1 1
−1/2
2 4
= −6
3 X2 3 1 0 0 0
5 3 2
− − −
2 4 5
0 S3 1 0 8 0 1
3 3
−
2 4
Zj 3 0 0 0
Zj-Cj -1/2 0 2 0 3/4 0
=4 =1 =0 = 2/5
= =0
Pivotedoperations
Cj -1 3 -2 0 0 0
Value of
Basic
basic X1 X2 X3 S1 S2 S3
variable
variable
5 X1 4 1 0 2/5 2/5 1/10 0
3 X2 5 0 1 3/5 1/5 3/10 0
35
0 S3 11 0 0 11 1 -1/2 1
Zj 5 3 33/5 13/5 7/5 0
Zj- Cj 6 0 23/5 13/5 7/5 0
X1 =4, x2 = 5 x3 = 0
Z = -4 +3 x 5 + 2 x 0 = 11
X1= 4, X2 = 5 ,X3 = 0
A minimization problem can be converted to maximization problem by change the sign of the
coefficients in the objective function.
Artificial variables
Since the coefficient of the surplus variables are negative this will not provide an initial base
feasible solution. This difficulty can be overcome by introducing another set of variables as
artificial variables with positive coefficients. These artificial variables have no value.
Hence the artificial variables are assigned very large positive coefficients or cost
represented by Min the objective function.
-2x1+ x2 – 5x3 ≥ -6
4x1+ x2 + x3 ≤6
36
Solution x1, x2, x3 >0
Subject to constraint
4x1 + x2 + x3 + 0S1+0S2+S3 = 6
S1 = 2, A1 = 6, S3 = 6
Ratio
1 2 1 0 0 0 -M
Value of
Basic
basic X1 X2 X3 S1 S2 S3 A1
variable
variable
2/1=2
0 S1 2 2 1 -1 1 0 0 0
6/1=6
-M A1 -6 -2 1 -5 0 -1 0 1
6/1=6
0 S3 6 4 1 1 0 0 1 0
Zj 2M -M 5M 0 0 0 0
Ratio
1 2 1 0 0 0 -M
Value
Basic
Cj of basic X1 X2 X3 S1 S2 S3 A1
variable
variable
2/0=
2 X2 2 2 +1 -1 1 0 0 2
infinite
4/1=4
-M A1 -8 -4 0 -4 0 -1 0 1
37
9/0=
0 S3 4 2 0 2 -1 0 1 0
Infinite
Zj 4+4m 2 4m-2 2 M 0 -M
Pivotedoperations
-6 = -6(2 x 1) = -4 6 = 6-(2 x 1) = 4
-6 = -6-(1x 1) = -4 4 =4-(2 x 1) = 2
1 = 1- (1 x1) = 0 1= 1- (-1 x 1) = 0
-5 = -5(-1 x 1 = 4) 1 = 1-(-1 x 1) = 2
0 = 0 – (0 x 1) = 0 0 = 0-(1 x 1) = -1
-1 = -1 – (0 x 0) =- 1 0= 0-(ox1)=0
0=0-(0x1)=0 1=1-(ox1)=1
1=1-(0x1)=1 0=0-(ox0)=0
Since alldijs are positive ,the solution is optimum
Z = 0 + 2 x 2 + 0=4
Problem 5
Subject to constraints
X1 + x2 = 200
X1 ≤ 80 , x2 ≥ 60
X1 ≥ 0 x2 ≥ 0
Solution
Subject to x1 + x2 + A1 = 200
X1 + S1 = 80
X2 – S2 + A2 = 60
38
This is a minimization case. Convert this to maximization problem by changing the sign of
the objective function
Z= -3 X1-8X2+0S1+0S2 –MA1-MA2
1 2 1 0 0 -M
Value of rATIO
Basic
Cj basic X1 X2 S1 S1 A1 A2
variable
variable
200/1=200
A1 -M 200 1 1 0 0 1 0
80/O=
X1 0 80 1 0 1 0 0 0
60/1=60
A2 -M 60 0 1 0 -1 0 1
Zj -M -2M 0 M -M M
Cj -3 -8 0 0 -M -M
Zj-Cj --M+3 - 0 M 0 0
2M+8
Highest negative ∆j is -2M + 8 .X2 is the incoming vector and A2 is the outgoing vector. Key
element is 1.
1 2 1 0 0 -M
Value of
Basic
Cj basic X1 X2 X3 S1 S2 A1
variable
variable
A1 -M 140 1 0 0 1 1
S1 0 80 1 0 1 0 0
X2 -8 60 0 1 0 -1 0
Zj -M -8 0 -M+8 -M
Cj -3 -8 0 0 -M
Zj-Cj --M+3 0 0 -M+8 0
39
Highest negative ∆j is –M+3. So X1 is the incoming vector. Minimum ratio is 80. So
outgoing vector is S1. Key element is 1
B CB XB X1 X2 S1 S2 A1 Ratio
A1 -M 60 0 0 -1 1 1 60
X1 -3 80 1 0 1 0 0 α
X2 -8 60 0 1 0 -1 0 -ve
Zj -3 -8 M-3 -M+8 -M
Cj -3 -8 0 0 -M
∆j 0 0 M-3 -M+8 0
S2 is the incoming vector. A1 is the outgoing vector. Key element is 1.
B CB XB X1 X2 S1 S2
S2 0 60 0 0 -1 1
X1 -3 80 1 0 1 0
X2 -8 120 0 1 1 0
Zj -3 -8 5 0 0
Cj -3 -8 0 0 0
∆j 0 0 +5 0
Since all∆js are positive, the solution is optimum
X1 = 80 , x2 = 120
Problem 6
40
Solution :Introduce Slack and Artificial variables
MaximiseZ = 2x1+x2+ 3x3+ 0S1 – MA1
Subject to constraintsX1 + x2 + 2x3 + S1 = 5
2x1+ 3x2 + 4x3+ A1 = 12
Put x1, x2, x3 = 0 S1=5, A1 = 12
0 2 1 3 0 -M Ratio
5 X1 X2 X3 S1 A1
5/2 =
0 S1 5 1 1 2 1 0
2.5
12/4 =
-M A1 12 2 3 4 0 1
3
Zj -2M -3M -4M 0 -M
2 1 3 0 -M
Coefficient
Basic Value of
of basic X1 X2 X3 S1 A1
variable basic value
variable
3 X3 2.5 0.5 0.5 1 0.5 0
- M A1 2 0 1 0 -2 1
Key row
IInd row
12 = 12 – (2.5 x 4) = 2
41
2 = 2-(0.5 x 4) = 0
3 = 3 – (0.5 x 4) = 1
4 = 4-(1 x 4) = 0
0= 0-(0.5 x 4) = -2
1= 1- (0 x 4) = 1
2 1 3 0 -M
Coefficient
Basic Value of
of basic X1 X2 X3 S1 A1
variable basic value
variable
3 X3 2.5 0.5 0.5 1 0.5 0
-M A 2 0 1.0 0 -2 1
Largest negative value is -1. Smallest positive ratio is 2. Hence A1 leaves and x2 enters
2 1 3 0
Coefficient of Basic Value of
X1 X2 X3 S1
basic variable variable basic value
3 X3 1.5 0.5 0 1 1.5
1 X2 2 0 1 0 -2
Key ratio=2/1=2, 0/1=0,1/1=1,0/1=0,2/-1=-2
Revised row
2.5=2.5-(2x0.5)=1.5
0.5=0.5-(0X0.5)=0.50
1=1-(OX0.5)=1
0.50=0.50-(1X0.5)=0
0.5=0.5-(2X0.5)=1.50
42
2 1 3 0 -M
Coefficient Value of
Basic
of basic basic X1 X2 X3 S1 Ratio
variable
variable value
1.5/0.5 =
3 X3 1.5 0.5 0 1 1.5
3
1 X2 2 0 1 0 -2 2/0 = α
Zj 1.5 1 3 2.5
Zj-Cj -0.5 0 0 2.5
2 1 3 0
Coefficient Value of
Basic
of basic basic X1 X2 X3 S1
variable
variable value
2 X1 3 1 0 2 3
1 X2 2 0 1 0 -2
Zj 2 1 4 4
Zj-Cj 0 0 1 4
Key row
1.5/0.5=3 ,0.5/0.5=1,0/0.5=0,1/0.5=2,1.5/0.5=3
Pivotted Operations
2 = 2 –(3 x 0) = 2
0 = 0-(1 x 0) = 0
1 = 1- (0 x 0) = 0
0 = 0 – (2 x 0) = 0
-2 = -2(3 x 0) = -2
43
Problem 7
X1 + 2x2 + x3 ≤ 430
X1+4x2 ≤ 420
X1 >0 x2 ≥ 0 x3 ≥ 0
Solution
Subject to constraints
Simplex table
3 2 5 0 0 0
Value Ratio
Coefficient
Basic of
of basic X1 X2 X3 S1 S2 S3
variable basic
variable
value
430/1=430
0 S1 430 1 2 1 1 0 0
460/2=230
0 S2 460 3 0 2 0 1 0
420/0=infinite
0 S3 420 1 4 0 0 0 1
Zj 0 0 0 0 0 0
Zj-Cj -3 -2 -5 0 0 0
Key row elements
½ = 0.5 , 0/2 = 0
Pivotted operation
44
3 2 5 0 0 0
Coefficient Value
Basic Exit
of basic of basic X1 X2 X3 S1 S2 S3
variable ratio
variable value
200/2 =
0 S1 200 -0.5 2 0 1 -0.5 0
100
=230/0=
5 X3 230 1.5 0 1 0 0.5 0
α
420/4 =
0 S3 420 1 4 0 0 0 1
105
Zj 7.5 0 5 0 2.5 0
Key row
430 = 430- (230 x 1) = 200
1=1-(1.5 x 1) = 2
2=2-(0 x 1) =2
1= 1- (1 x 1) = 0
0.5=0.5-(0.5x1)=-0.5
0=0-(0x1)=0
Pivotted operations
420 = 420 – (230 x 0) = 420
1= 1- (1.5 x 0) = 4
0=0-(1 x 0)=0
0=0-(1x0) =0
0=0-(0.5 x 0) = 0
1=1-(0x0) = 1
All dijs are not positive. Hence the solution is not optimal
Largest negative value is -2
45
3 2 5 0 0 0
Coefficient Value of
Basic
of basic basic X1 X2 X3 S1 S2 S3
variable
variable value
2 X2 100 -0.25 1 0 0.5 -0.25 0
0 S3 20 2 0 0 -2 1 1
Key row
2nd row
0= 0- (1 x 0) = 0
1 = 1- (0 x 0) = 1
0= 0- (0.5 x 0) = 0
0 = 0-(0 x 0) = 0
IIIrd row
1 = 1-(-0.25 x 4) = 2
4 = 4- (1 x 4) = 0
0 = 0-(0 x 4) = 0
0=0-(0.5 x 4) = -2
0=0-(-0.25 x 4) = 1
1=1-(0x4) =1
X3 = 230
X1 = 0
1. Infeasible solution
Here one or more artificial variables remain in the final solution with non zero values
2. Unbounded problem
An LPP is said to be unbounded solution, if the objective function can be increased or
decreased arbitrarily, then is no finite optimum of the objective function. Here no exit
variable exists since the simplex table contain only negative ratios.
3. Problems with multiple optimum
There can be more than one optimal solution for certain problem. In such cases, the
problem does not have a single or unique solution. In the final simplex table, if the
index row indicates the value for any non basic variable to zero it means that alternate
optimal solution exists for the problem. Thus the problem has an alternate solution
which is optimal.
A basic feasible solution of LPP is said to be degenerate, if at least one of the basic variable is
zero. In an LPP, degeneracy occurs at the following stages.
1. Degeneracy may appear at the very firstiteration if one of the basic variable is zero.
2. Degeneracy if may occur at some subsequent iterations This happens when there is a
tie for selecting exit variable
When the simplest algorithm is applied to a degenerate feasible solution to get a new BFs,
the value of the objective function may remain unchanged.
In some cases, the problem of cycling may occurs, the same sequence of simplex table are
repeated for ever without reaching optimum solution. In this case
47
ofcycling,variableswhich have left the basic solution in the previous iteration reenters and
no optimal solution will be reached.
Duality theorem sates that for every maximization (minimization) problem in linear
programming there is a unique similar problem of minimization (or maximization)
involving same data which describe the original problem
Applications of dual
1. If the primal problem contains a large number of rows (constraints) and smaller
number of columns (variables) computational procedures can be considerably reduced
by converting it into dual
2. Calculation of dual checks the accuracy of the primal solution
3. Disability in linear programme has certain far reaching consequence of economic
nature
4. It gives additional information as to how the optimal solution changes as a result of
changes in the coefficient and formulation of the problem. This is termed post
optimality or sensitivity analysis
Sensitivity analysis
Analysis of such problems assessing due to the slight change made in the parameter or
structure of a given LPP after attaining its optimum solution known as
The sensitivity analysis is concerned with the extent of optimal solution of LPP for
changes in one or more
48
1. Coefficient in the objective function
2. Coefficient of variables in the constraints
3. RHS constraints in the constraints
The sensitivity analysis makes the LPP solution to a dynamic tool that evaluates change
conditions.
The following post optimally problems can be answered with the help of sensitivity
analysis
Advantages of LPP
Subject to constraints
x1 + x2+ x3 ≤ 5
2x1 + x3 ≤ 10
X2 + 3x3 ≤ 15
X1, x2 , x3 > 0
Solution
Dual
Y1 + y3 ≥ 1
Y1 + y2+ 3y3 ≥ 2
49
Problem 9
8x1+ 5x2 ≥ 60
3x1+ 4x2 = 50
X1, x2 ≥ 0
Solution
Minimize
Subject to constraints
-8X1-5X2 ≤ 60
(Third constraints equation has to be converted unit two inequalities in opposite direction)
3x1+ 4x2 ≤ 50
3x1+ 4x2 ≥ 50
-3x1-4x2 ≤ -50
Problem 10
Minimise Z= 2x2+5X3
Subject to
50
X1+X2>2
2X1+X2+6X3<6
X1-X2+3X3=4
X1,X2,X3>0
-2x1-x2-6x3> -6
X1-x2+3x3<4
X1-x2+3x3>4
-x1+ x2 – 3x3 ≥ -4
Subject to
W1-2W2-W3 + W4 ≤ 0
W1 –W2+ W3 – W4 ≤ 2
W1,w2,w3>0
Activity Question
3x1 + x2 + x3 > 23
7X1 – x2 = 4
X1, x2, x3 ≥ 0
51
Problem 12
Subject to x1 +x2 + x3 = 22
X1 + x2 ≥ 0 x1, x2, x3 ≥ 0
Problem 13
2x1+3X2 +x3 ≥ 4
X1, x2, x3 ≥ 0
Subject to constraints
2w1+ 2w2 ≤ 4
W1+ 3w2 ≤ 2
4w1+ w2 ≤3
W1,w2 ≥ 0
52
5 4 0 0 0
Coefficient Value of Ratio
Basic
of basic basic w1 w2 S1 S2 S3
variable
variable value
4/2=2
0 S1 4 2 2 1 0 0
2/0
0 S2 2 0 3 0 1 0
¾=0.75
0 S3 3 4 1 0 0 1
Zj 0 0 0 0 0
Zj-Cj -5 -4 0 0 0 0
S3 leaves w1 enters
5 4 0 0 0
Coefficien Value of
Basic
t of basic basic w1 w2 S1 S2 S3
variable
variable value
2.5/1.5 = 2
0 S1 2.5 0 1.5 1 0 0
2/3=0.67
0 S2 2 0 3 0 1 0
0.75/0.25 = 3
5 W1 0.75 1 0.25 0 0 0.25
Zj +5 +1.25 0 0 1.25
Pivotted operation
2 = 2 – (1 x2) = 0
2 = 2- (0 .25 x 2 ) = 1.5
1= 1- (0x 2) = 1
53
0= 0-(0x2) = 0
0= 0-(0x 2) = 0
2nd row
2 = 2- (0.75 x 0) = 2
0= 0- (1x 0)= 0
3= 3-(0.25 x 0) = 0
0=0-(oxo)=0
1 = 1-(0 x 0) = 1
0 = 0-(0 x0) = 0
5 4 0 0 0
Coefficient Value of
Basic
of basic basic w1 w2 S1 S2 S3
variable
variable value
0 S1 1.5 0 0 1 -0.405 -0.5
4 W2 0.67 0 1 0 0.33 0
5 W1 0.5825 1 0 0 -0.0825 0
1st row
1= 1 – (1.5 x 0) = 1
54
0 = 0- (1.5 x 0) = 0
111rd row
1= 1-(0.25 x 0) = 1
0 = 0- (0.25 x 0) = 0
5 4 0 0 0
Coefficient Value of
Basic
of basic basic w1 w2 S1 S2 S3
variable
variable value
0 S1 1.5 0 0 1 -0.495 -0.5
4 W2 0.67 0 1 0 0.33 0
Zj 5 4 0 0.9075 1.25
Since all dijs are positive and zero, the solution is optimal
W1 = 0.5825 w2 = 0.67
Activity Question 14
Subject to constraints
55
2x1 + x2 + x3 ≥30
X1 + x2 + x3 ≥ 20
X1+ 2x2 + x3 ≥ 20
Problem 15
Subject to 2x1 + x2 ≤ 1
X1 + 4x2 ≥ 6, x1, x2 ≥ 0
Problem 16
Problem 17
subject to constraints
2x1+ 6x2 + x3 ≥ 7
X1- 2x2 + x3 ≤ 5
X1 x2, x3 ≥ 0
Solution
2x1 + 6x2 + x3 ≥ 7
X1 – 2x2+ x3 ≤ 5
56
Introduce stack artificial variables
X1-2x2+ x3 +S3 = 5
S1 = 8, A1 = 7, S3 = 5
5 4 0 0 0
Coefficient Value
Basic
of basic of basic X1 X2 X3 S1 S2 S3 A1 Ratio
variable
variable value
S1=0 S1 8 3 4 -5 1 0 0 0 8/4=2
-M A1 7 2 6 1 0 -1 0 1 7/6=1.17
5/-2=
S3=0 S3 5 1 -2 1 0 0 1 0
-2.5
Zj -2M -6M -M 0 M 0 -M
Key row
Pivotted row
57
0=0-(0.17x4)= - 0.68
0=0-(0.17x2)= 0.34
5 4 0 0 0 0 -M
Coefficie
Basic
nt of Value of
variabl X1 X2 X3 S1 S2 S3 A1 R
basic basic value
e
variable
0 S1 3.32 1.68 0 -5.68 1 0.68 0 -0.68
All dijs are positive. Hence solution is optimal. Optimum value of objective function
where x2 = 1.17 ,X1 = 0 x3 = 0
3 x 0 + 5 x 1.7 + 0
= 5.85
58
Unit II
Assignment Problem
Assignment problems deals with the allocation of jobs to person or allocation of tasks to
facilities, the assignment is to be done on the basics of one job to one person or one task
to one facility
Cost matrix
n x nmatrix called cost matrix where Cij is the cost of assigning ith source to jth
destination
Source 1 2 3 ….. n
2 C21 C22……C2n
Z=∑ ∑ Cijxij
Subject to condition ∑!" = 1 for j = 1, 2 ….. n for which one job is to be done
by ith person where i=1 ….. n
∑!" = 1fori = 1,2 … .n which means one person to be assigned the jth job
MethodofsolvingAssignment problem
(Hungarian Alogarithm or Reduced Matrix Method)
1. Subtract the smallest element of each row in the cost matrix from every element of
that row
2. Subtract the smallest element of each column in the reduced matrix from every
element of that column
3. a.Starting with rows of the matrix obtained, examine all the rows having exactly one
zero element. Enclose this zero within showing that assignment is made there.
59
Cross and all other zeros in the column to show that they cannot be sued to make
other assignments. Proceed in this way until last row is examined
b.Examine all the columns with one unmarked zero. Marked at the zero and cross
all the zeros of the row in which is marked. Proceed in this way till the last column
is examined
c.Continue these operations (a) and(b) successively until we reach any of the
following solution
d. If all the zeros are enclosed by in this case we have maximal assignment and
assignment in every row or column or got solution.
2. Remaining unmarked zeros lie at least two rows or column. It does not contain
assignment in all rows and all columns
In this case the following procedure is followed
4. Draw the minimum number of horizontal and vertical lines necessary to cover all
zeros at least once.
5. Select the smallest of the elements which is not covered by lines.subtract it from all
the elements that do not have a line through them, add it to the every element that lies
at the interaction of two lines and leave the remaining elements of the matrix
unchanged.
6. Now reapply steps 3 to 5 to the modified matrix
60
2 7 11 10 14 12
3 5 6 9 12 14
4 13 15 11 10 7
Allocation
8 9 2 [0] 6
[0] 4 3 7 5
0 1 4 7 9
6 8 4 3 [0]
0 [0] 0 0 0
8 9 2 0 6
0 4 3 7 5
0 1 4 7 9
6 8 4 3 0
0 0 0 0 0
61
8 8 1 0 7
0 3 2 6 5
0 0 3 6 9
6 7 3 2 0
1 0 0 0 1
8 8 1 [0] 7
[0] 3 2 6 5
0 [0] 3 6 9
6 7 3 2 [0]
1 0 [0] 0 1
Optimum solution
Job 1 – Machine 4-2
Job II – machine 1 – 7
Job III – Machine 2-6
Job IV – Machine 5-7
Hence total optimal cost = 2 + 7 + 6 + 7 = Rs. 22
Problem 2 ( UNI Feb 2016)Marks 20 Marks
A company has four machines to do three jobs. Each job can be assigned to one and
only one machine. The cost of each job on each machine is given in the following
table.
What are the job assignments which will minimize the cost?
Solutions
Machines
1 2 3 4
A 18 24 28 32
B 8 13 17 19
C 10 15 19 22
62
Now it becomes a balanced assignment problem
0 6 10 14
0 5 9 11
0 5 9 12
0 0 0 0
[0] 6 10 14
0 5 9 11
0 5 9 12
0 [0] 0 0
Draw minimum number of vertical and horizontal lines to cover all zeros
0 6 10 14
0 5 9 11
0 5 9 12
0 0 0 0
[0] 1 5 9
0 0 4 6
0 0 4 7
5 0 [0] 0
3 rdrow allocation not made. Draw minimum number of horizontal and vertical
lines to cover all zeros.
63
0 1 5 9
0 0 4 6
0 0 4 7
5 0 0 0
Least element uncovered is 4. Subtract 4 with respective elements and add 4 at
intersection points.
0 1 1 5
0 0 0 2
0 0 0 3
9 4 0 0
[0] 1 1 5
0 [0] 0 5
0 0 [0] 3
9 4 0 [0]
0 1 1 5
0 0 0 2
0 0 0 3
9 4 0 0
A- 1 = 18 + 13+ 19
B- 2 = Rs. 50
C- 3
D- 4
64
A B C D
I 1 4 6 3
II 9 7 10 9
III 4 5 11 7
IV 8 7 8 5
Solution
0 3 5 2
2 0 3 2
0 1 7 3
3 2 3 0
0 3 2 2
2 0 0 2
0 1 4 3
3 2 0 0
[0] 3 2 2
2 [0] 0 2
0 1 4 3
3 2 [0] [0]
0 3 2 2
2 0 0 2
0 1 4 3
3 2 0 0
Subtracting 1 from elements not marked and add 1to intersection points
65
0 2 1 1
2 0 0 3
0 0 3 3
4 2 0 1
[0] 2 1 1
2 0 [0] 3
0 [0] 3 3
4 2 0 1
I–1
II- 3
III- 2
IV – 4
Optimum cost
= 1+ 10 + 5 + 5 = 21
Problem 4
Five different machines can do any of the five required jobs with different profits
resulting from each assignment as given below
A B C D E
1 30 37 40 28 40
2 40 24 27 21 36
3 40 32 33 30 35
4 25 38 40 36 36
5 29 62 41 34 39
Find out the maximum profit possible through optimal assignment
Solution
66
This is a maximization problem. So convert into minimization problem. Subtract all
elements from highest of elements (ie 62). After this problem gets converted to
minimization problem.
32 35 22 34 22
22 38 35 41 26
22 30 29 32 27
37 24 22 26 26
33 0 21 28 23
Subtracting smallest element from each row from every element of row and smallest
element from each column from every element.
10 3 0 12 [0]
[0] 16 13 10 4
0 8 7 10 5
15 2 0 4 4
33 [0] 21 24 23
10 3 0 8 0
0 16 13 15 4
0 8 7 6 5
15 2 0 0 4
33 0 21 24 23
Smallest element uncovered is 4 subtract 4 from elements and add 4to intersection
points
14 3 [0] 8 0
0 12 9 11 [0]
[0] 4 3 2 1
19 2 0 [0] 4
37 [0] 21 24 23
Assignment is complete
I- C – 40
II- E – 36
III- 1 – 40
67
IV- D – 36
V- B – 62
Travelling sales man problem is a special type of routing problem. The routing
problems are those where we have to select a route, from an origin to destination,
which yields minimum cost.
Suppose a salesman has to visit in cites .He wishes to start from a particular city visits
each city once and then return to his starting time. Objective is to select the sequence
in which cities are visited in such a way that his total travelling time is minimized. If
there are 4 cities A,B, C, D then once solution can to A to C, C to D, D to B, B to A
Travelling sales man problem is very similar to assignment problem except that in the
former there is an additionalrestriction . The addition restriction is choosing a
sequence which can minimize cost. This is the route condition.
Problem 5
A sales man has to visit fivecities A, B, C, D, E. The between cities are given in the
table below. If the sales man streets from city A and has to comeback to city A which
route should be select so that the distance travelled by him is minimized.
A B C D E
A - 4 7 3 4
B 4 + 6 3 4
C 7 6 - 7 5
D 3 3 6 - 7
E 4 4 5 7 -
Solution
68
A B C D E
A α 4 7 3 4
B 4 Α 6 3 4
C 7 6 α 7 5
D 3 3 7 α 7
E 4 4 5 7 α
Making row and column subtraction and making zero assignment we have
A B C D E
A α 1 4 [0] 1
B 1 α 3 0 1
C 2 1 α 2 [0]
D [0] 0 4 α 4
E 0 [0] 1 3 α
Α 1 4 0 1
1 Α 2 0 1
2 1 α 2 0
0 0 4 α 4
0 0 0 2 α
Least element uncovered by lines is 1. Subtracting or adding 1 to the respective elements and
making zero assignment
Α 0 3 0 0
0 Α 1 0 0
2 1 α 3 0
0 0 3 α 4
0 0 0 3 α
Draw minimum number of vertical lines and horizontal lines to cover all zeroes
Α 0 2 0 0
0 Α 1 0 0
2 1 Α 3 0
0 0 3 α 4
0 0 0 3 α
69
Draw minimum no. of horizontal vertical lines
α 0 2 0 0
0 Α 1 0 0
2 1 α 3 0
0 0 3 α 4
0 0 0 3 α
Draw
Α [0] 2 [0] 0
0 Α [1] [0] α
2 1 α 3 [0]
[0] [0] 3 α α
[0] 0 0 3 α
Minimum distance travelled
A – D, B – C, C- E, D – B, E- A
A–D–B–C–E–A
Activity Questions
Problem 6
From city 1 2 3 4 5
1 - 10 25 25 10
2 1 - 10 15 2
3 8 9 - 20 10
4 14 10 24 - 15
5 10 8 25 27 -
Problem 7
Given below is a matrix showing profit for different jobs done through differentmachines .
Find an assignment programme which will maximize the total profit
70
M1 M2 M3 M4
J1 51 53 54 50
J2 47 50 48 50
J3 49 50 60 61
J4 63 64 60 61
Problem 8
I II III IV
A 32 26 35 38
B 27 24 26 32
C 28 22 25 34
D 10 10 16 16
Problem 9
A company is faced with the problem of assigning six different machines to five different
jobs. The costs are estimated as follows (in hundreds of rupees)
jobs
machines A B C D E
1 5 10 2 12 2
2 4 10 3 14 6
3 6 13 5 16 9
4 7 14 4 18 12
5 8 14 6 18 12
6 12 18 10 20 12
Solve the problem assuming that objective is to maximize total cost
Problem 10
The owner of a small machine shop has four machines available to assign job for the day.
Five jobs are affected with expected profit fore each machine on each job which are as
follows.
jobs
Machine A B C D E
1 62 78 50 111 82
2 71 84 61 73 59
3 87 92 111 71 81
4 48 64 87 77 80
71
Find by using assignment method, the assignment of machines to job that will result in a
maximum profit which job should be declined
Machines
M4 M5
Job M1 M2 M3
3 9
A 7 5 2
4 5 10
B 6 5
6 8
C 5 4 5
2 6
D 8 3 3
Solution
7 5 2 3 9
6 5 4 5 10
5 4 5 6 8
8 3 3 2 6
0 0 0 0 0
72
Subtract the lowest element from each row
1 7
5 3 0
0 1 6
2 1
2 4
1 0 1
0 4
6 1 1
0 0
0 0 0
5 3 0 1 7
2 1 0 1 6
1 0 1 2 4
6 1 1 0 4
0 0 0 0 0
Allocation
5 3 [0] 1 7
2 1 0 1 6
1 [0] 1 2 4
6 1 1 [0] 4
[0] 0 0 0 0
All the rows does not have assignment
5 3 0 1 7
2 1 0 1 6
1 0 1 2 4
6 1 1 0 4
0 0 0 0 0
73
Lowest element not covered by lines is 1 .Subtract the lowest element from other elements
and add 1 to the intersection points
4 3 0 0 6
1 3 0 0 5
0 0 1 2 3
5 1 1 0 3
0 1 1 1 0
Allocation
4 3 0 0 6
1 3 0 0 5
0 [0] 1 2 3
5 1 1 [0] 3
0 1 1 1 [0]
Draw Minimum no of horizontal and vertical lines to cover zeroes
4 3 0 0 6
1 3 0 0 5
0 0 1 2 3
5 1 1 0 3
0 1 1 1 0
3 2 0 0 6
0 2 1 0 6
1 0 2 1 2
6 1 1 0 4
0 1 1 1 1
3 2 [0] 0 6
0 2 1 0 6
1 [0] 2 1 2
6 1 1 [0] 4
[0] 1 1 1 1
Allocation
I- M3 – 2
II- M5 – 10
III- M2 – 4
74
IV- M4 – 2
V- M1 – 0
Optimum Allocation = 2+ 10 + 4 + 2 +0 = 18
TRANSPORTATION PROBLEM
Transportation techniques can be applied not only to cost minimizing problem but also to
time minimizing problem, profit maximizing problem.
1. To minimize the transportation costs from factories to ware house or ware houses to
markets.
2. To determine the lowest cost allocation for a new factory
3. To determine minimum cost production schedule.
Transportation table
D1 D2 ………Dn Availability
Requirement b1 b2 ……….. bn
The matrix is known as transportation table. Origin is denoted as 01, 02….. On and
destination is noted as D1, D2,….Dn and Cj for the cost associated with transportation
problem.
75
Basic assumptions in transportation techniques
Definitions
Therefore for feasibility the sum of allocation in the rows must be equal to availability in
that row.
Similarly sum of allocations in the column must be equal to demand in that column
Initial feasible solution are those which satisfy the rim requirement. That is the allocation
made in every row taken together is equal to availability shown in that row similarly for each
column, the total allocation should be equal to the requirement in that column
76
3. Vogel approximation method is more preferred since the initial basic feasible solution
obtained either optimal or very close to the optimal solution.
Alogarithm steps
Step 1 : Allocate to the cell(1,1) maximum possible amount which is the minimum of row
total and column total. So either arow or column total gets exhausted. So cross of total
row or column as the case may be
Step 2 : Consider the reduced matrix. In that matrix, allocate to the cell (1,1) maximum
possible amount (Which is the minimum of present row total or column total)
Step 3 : Repeat the above steps until all the available quantities are exhausted.
Problem 1
Find the initial feasible solution to the transportation problem by northwest corner rule
Origin D1 D2 D3 Supply
01 2 7 4 5
02 3 3 1 8
03 5 4 7 7
04 1 6 2 14
Demand 7 9 18
Solution
5 2 7 4 5
3 3 1 8
5 4 7 7
1 6 2 14
7 9 18
2 3 3 1 8
5 4 7 7
1 6 2 14
2 9 18
77
6 3 1 6
4 7 7
6 2 14
9 18
3
4 7 7
6 2 14
3 18
4 7 4
14 14
2
18
Choose the cell having lowest cost in the matrix allocate there, as much possible which is
the minimum of row total and column total. This either a row total or column total is
exhausted.
Cross off the corresponding row or column from the reduced matrix, locate the cell
having lowest cost. Allocate to that cell maximum possible thus leading to a feasible
reduced matrix.
78
Problem 2
Find the initial basic feasible solution to the following transportation problem by lowest
cost entry method
W1 W2 W3
F1 2 7 4 5
F2 3 3 1 8
F3 5 4 7 7
F4 1 6 2 14
7 9 18
Solution
Lowest cost on cells (2,3) & (4,1) select one of these say (2,3). Allocate minimum of
8and 18 ie.8. Thus F2 is exhausted
2 7 4 5
8
3 3 1 1
5 4 7 7
1 6 2 14
7 9 10
2 7 4 5
5 4 7 7
7
1 6 2 7
9 10
7 4 5
4 7 7
6 7 2 7
9 10
Minimum cost
3 4
7 5
4 7 7
9 3
79
2
7 2 7 2
7
4 7
9 9
2 3
2 7 4 5
8
3 3 1 8
7 4
5 7 7
7 7
1 6 2
7 9 18
Total transportation cost
Step 1
Since this method, we write the difference between smallest and second smallest costs in
each column below the corresponding columns, within brackets.similarly write similar
difference in each row to the weight of corresponding row. These differences are known
as penalty
Step 2
Select the row or column having the largest penalty and allocate the maximum possible
amount to the cell with the lowest cost in that row or column as the case may be. This
either the row total is column total is completely exhausted (cross of that row or column).
Construct thereduced matrix with the remaining rows and column
Step 3 : For the reduced matrix obtained apply step 1 & 2 until all rows and column totals
are exhausted. The initial solution obtained by Vogelsmethod is more close to the optimal
solution than the solution obtained by other two method
Problem 3
80
W1 W2` W3 Supply
F1 2 7 4 5
F2 3 3 1 8
F3 5 4 7 7
F4 1 6 2 14
W1 W2` W3 Supply
5
F1 2 7 4 5(2)
F2 3 3 1 8(2)
F3 5 4 7 7(1)
F4 1 6 2 14(1)
Demand
7 9 18
(1) (1) (1)
Penalties
For rows = 4- 2 = 2, 3-1 = 2, 5-4 = 1, 2-1 = 1
For columns 2-1 = 1, 2-1 = 1
Max penalty is 2 (1st row)
8
3 3 1 8(2)
5 4 7 7(1)
1 6 2 14(1)
2 9 18
(2) (1) (1)
Max penalty is in row 2
5 4 7 7(1)
1 6 2 14(1)
2 (4) 9(2) 10(5)
Maximum penalty is 5
5 4 7 7(1)
10
1 6 2 14(1)
2 9 10
81
5
4 7(1)
2
1 6 4(5)
2(4) 9(2)
7
4 7
6 2
9(2)
2
6 2
9
Max penalty is 5
Total cost = (5 x2) + (8x1) +(7 x4) +(2 x ) (2 x6) + (10x2) = Rs. 80
Problem 4
Solve the transportation problem(UNI Feb 2015)
Distribution centre
Origin 1 2 3 4 Availability
I 21 16 25 13 11
II 17 18 14 23 13
III 32 27 18 41 19
Req 6 10 12 15 43
Solution
Availability = 43
Requirement = 43
We will find the initial basic feasible solution using VAM
1 2 3 4
21 16 25 13 11(3)
17 18 14 23 13(3)
32 27 18 41 19(9)
6 10 12 15
(4) (2) (4) (10)
82
Max penalty is 10 (4th column)
11
21 16 25 13 11
17 18 14 23 13
32 27 18 41 19
6 10 12 15
17 18 14 23 13(3)
32 27 18 41 19(9)
6 10 12 4
(15) (9) (4) (18)
Max penalty is 4th column
4
23
17 18 14 13(3)
32 27 18 41 19(9)
6 10 12 4
Availability
17 18 14 9
32 27 18 19
6 10 12
6
17 18 14 9(3)
32 27 18 19(9)
6 10 12
(15) (9) (4)
Maximum penalty is 1st column
3
18 14 3(4)
27 18 19(9)
10(9) 12(4)
7 12 18
27 19(9)
7 12
83
Initial basic feasible solution
(11x 13) + (6 x 7) + (3 x 18) + (4x23) + (7 x 27) + (12 x 8) = Rs. 796
Origin A B C D Availability
P 11 13 17 14 250
Q 16 18 14 10 300
R 21 24 13 10 400
Req 200 225 275 250
Solution
11 13 17 14 250(2)
16 18 14 10 300(4)
21 24 13 10 400(3)
200 225 275 250
(5) (5) (1) (4)
13 17 14 50
18 14 10 300
24 13 10 400
225 275 250
13 17 14 50(1)
18 14 10 300(4)
24 13 10 400(3)
84
225 275 250
(5) (1) (4)
50
13 17 14 50(1)
18 14 10 300(4)
18 14 10 300(4)
24 13 10 400(2)
225 275 250
(5) (1) (4)
17518
14 10 300(4)
24 13 10 400(3)
175 275 250
(6) (1) (0)
125 125(4)
14 10
13 10 400(3)
275 250
(1) (0)
275 125
150
13
200 50
11 13 17 14 250
175 125
16 18 14 10 300
275
21 24 13 10 400
200 225 275 250
Total transportation cost = (1 x 200) + (13 x 50) + (18 x 175) + (10x 125) + (13 x 275) +
(10 x 125)
200 + 650 + 3150 + 1250 + 3575 + 1250 = Rs. 7240
Optimal Solution
85
By applying Vogelsmethod or lowest cost entry method or North west corner rule, an initial
basic feasible solution can be obtained. Next step is to examine whether the solution can be
improved. For the we have to conduct the test of optimality by modified distribution method
(MODI’s method)
MODIs Method
Step 1 : When the initial basic feasible solution obtained, some cells are occupied and other
unoccupied. No of occupied cells is (m+n-1) let cij be the cost of cell (i,j)
Then we determine (m+n) number called Ui&Vj values by framing (m+n) numbers called
Ui&Vj values by framing (m+n-1) equations of the form Ui+Vj = Cij correspond to each of
the occupied cell.
For solving the equations, we take one of Ui and Vj values as zero (since no of unknown is
one more than the number of equation)
Step 2 : Then we calculate cell evaluation known as dijvalues for unoccupied cells by the
formula dj = Cj –(Ui+Vj)
Step 3 : If all the dj values are positive, the solution is optimal and unique. If at least one of
them is zero, and other positive, the solution is optimal but alternative solution exists. If at
least dij is negative, the solution is not optimal.
Step 4 : If the solution is not optimal, make reallocation. Give maximum allocation to the cell
for which dij is negative, making one of the occupied cells empty
Then steps 1 to 4 will be repeated until solution becomes optimal.
Problem 6 (UNI NOV 2012- 5 Marks)
Supply
21 16 25 13
11
17 18 14 23 13
32 27 18 41 19
6 10 12 15
Solution
First we have to find the initial basic feasible solution by vogels approximation method
Initial basic feasible solution
W1 V1 V2 V3 V4 Supply
11
W2 21 16 25 13 11
6 3 4
W3 17 18 14 23 13
7 12
W4 32 27 18 41 19
6 10 12 15
86
Minimum transportation cost
= (13x 11) + (17 x6) + (18 x3) + (23 x3) + (23 x 4) + (27 x 7) + (18 x12) = Rs. 796
Total no of allocation = 6
(m + n-1) = (4 + 3-1) = 6
Since total no of allocation = (m +n-1) the allocation are independent,non degenerate feasible
solution is established.
U1+ v4 = 13
U2 + V1 = 17
U2 + V2 = 18
U2 + V4 = 23
U3 + V2 = 27
U3 + V3 = 18
Put U2 =0 V1= 17 U2 = 18
U3 + 18 = 27 U3 = 9
9 + V3 = 18 V3 = 18-9 = 9
U1 + 23 = 13
U1 = -10
Cij
21 16 25 X
X X 14 X
32 X X 41
Ui+VJ)
87
7 8 -1 X
X X 9 X
26 X X 32
14 8 26 x
X X 5 X
6 X X 9
Problem 7
2 7 4 Supply 5
3 3 1 8
5 4 7 7
Demand 1 6 2 14
Solution
2 7 4 Supply 5
3 3 1 8
5 4 7 7
1 6 2 14
Demand 7 9 18
m+n-1= 4=3-1=6
Total no of allocation =6
U1 + V1 = 2
U2 + V3 = 1
88
U3 + V2 = 4
U4 + V2 = 6
U4 + V3 = 2
U4 occurs max no of times u4 = 0
V1 = 1, V2 = 6, V3 = 2
U2 + 2 = 1 U2 = -1
U3 = 4-6 = -2
U3 + V2 = 4
U3 + 6 = 4
U1+ 1 = 2, u1 =1
For unoccupied cells Cy – (Uc+ Vj)
Cij
X 7 4
3 3 X
5 X 7
X X x
Ui+Vj
X 1+6=7 1+2 = 3
-1+1=0 -1+6=5 X
-2+1 = -1 X -2+2=0
Dij= Cij – (ui1+ Vj)
X 0 4-3=1
3-0=3 3-5=-2 X
5+1-6 X 7-0=7
X X x
Since one of the dij values is negative the solution is not optimal. Make reallocation. Give
max allocation to the cell with negative dij occurs (2,2)
5 2 7 4 5
2 6
3 3 1 8
7 4
5 7 7
2 12
1 6 2 14
7 9 18
89
U4 + V1 = 1
U4+ V3 = 2
Put U4 = 0 V3 = 2 V1 = 1
Put U1 = 1 U2 = -1 V2 = 4 U3 = 0
Cij Ui + Vj
X 7 4 X 4+1=5 2+1=3
3 X X 1-1=0 X X
5 X 7 0+1=1 X 0+2 =2
X 6 X X 0+4 = 4 x
dij= Cij – Ui + Vj
X 2 1
3 X X
4 X 5
X 2 x
Now all dijsare positive. Solution is optimal
Source A B C D Supply
1 15 1 42 33 23
2 80 42 26 81 44
3 90 40 66 60 33
Demand 23 31 16 30 100
Solution
This is a maximization case problem. Highest value is 90. Subtracting all the elements by 90
to convert problem into minimization problem
1 75 89 48 57 23
2 10 48 64 9 44
3 0 50 24 30 33
90
Demand 23 31 16 30 100
75 89 48 57 23 (9)
10 48 64 9 44 (1)
0 50 24 30 33 (24)
23 31 16 30 100
(10) (2) (24) (21)
89 48 57 Supply 23
48 64 9 44
50 24 30 10
Demand 31 16 30
89 48 57 23(9)
48 64 30 9 44(39)
50 24 30 10(6)
31 16 30
(2) (24) (31)
Supply
89 48 23
48 64 14
50 24 10
Demand
supply
89 16 48 23(41)
48 64 14(16)
50 24 10(26)
31 16
(2) (24)
89 7
48 14
50 10
31
91
Supply
7 16
75 89 48 57 23
14 30
10 48 64 9 44
23 10
0 50 24 30 33
Demand 23 31 16 30 100
Maximum Profit = (89 x 7) + (48 + 14) + (10 x 50) + (23 x 0) + (48 x 1) + (9 x 30) = Rs.
2833
available amounts is not equal to sum of all requirements in destination together ∑ $ ≠ ∑ '
A transportation problem is said to be unbalanced transportation problem if the sum of all
1. Supply is less than demand : Lost sales &loosing customers. Introduce dummy origin
2. Supply is greater than demand
Profit is reduced by the loss caused by unsold quantity. Dummy destinations are
introduced to absorb the excess capacity of origin
An unbalanced transportation problem can be converted to a balanced transportation
problem by introducing a fictious source of destination which will provide the surplus
or demand. The cost of transporting unit from the fictiouss (source or destination) is
taken to be zero
After converting the unbalanced transport problem into a balanced transportation
problem by a dummy source or destination, it can be solved by Northwest corner
rule/ least cost method / VAM
Degeneracy in transportation problem
In a transportation problem, degeneracy occurs ,whenever the number of individual
allocation less than (m+n-1) where m & n are respectively number of rows and
columns of the transportationmatrix.
Degeneracy in the transportation problem are developed in 2 ways
1.The basic feasible solution may be degenerated from the initial stage onwards.
2. They may become degenerate at any immediate stage
In such cases, we allocate a small number ∆ which is equal to zero to one or more
empty cells so that the total no of allocations is (m+n-1)
Problem 9
Determine the optimal transportation plan from the following table giving the plant to
make shipping costs and quantities required at each market and available at each plant
Plant W1 W2 W3 W4 Available
F1 11 20 7 8 50
92
F2 21 16 10 12 40
F3 8 12 18 9 70
30 25 35 40
Solution
Here the total requirement of the market = 30+ 25+ 35 + 40 = 130
The total availability at the plants = 50 + 40 + 70 = 160 .
This is an unbalanced assignment problem since the total availability is 30 more than
the total requirements
So we convert this problem to a balance one by introducing a fictions market W5 with
requirement 30 such that the cost of transportation from plants to the market W5 are
zero.
Balanced transportation problem is
Plant W1 W2 W3 W4 W5 Available
F1 11 20 7 8 0 40
F2 21 16 10 12 0 70
F3 8 12 18 9 0
30 25 35 40 30 160
93
7 8 50 (1)
10 12 10 (2)
18 15
9 15 (9)
35 40
(3) (1)
7 25 8 50(1)
10 12 10(2)
35 25
(3) (4)
25 7 25
10 10
35
Plant W1 W2 W3 W4 W5 Available
25
F1 11 20 25 7 8 0 50
10 30
F2 21 16 10 12 0 40
30 25 15
F3 8 12 18 9 0 70
30 25 25 40 30 160
A B C Available
X 50 30 220 1
94
90 45 170 3
Z 250 200 50 4
Requirement 4 2 2
Shipping costs Rs. 10 per load mile what shipping should be used?
Solution
By vogel method, we get the following initial basic feasible solution
A B C
X 1 50 30 220 1
y 3 90 45 170 3
Z 250 2
200 2
50 4
4 2 2
Since the total allocations is 4 which is one less than the (m+n-1)=5,the solution is
degenerate
Now to resolve this degeneracy,y we allow a very small amount ∆ to the cell (2,2)
getting 5 allocations at independent position.
New solution is
A B C
X 1 50 30 220 1
y 3 90 ∆
45 170 3
Z 250 2
200 2
50 4
4 2 2
Testing for optimality
For occupied cells
U1+ V1 = 50
U2 + V1 = 90
U2 + V2 = 45
U3 + V2 = 200
U3 + V3 = 500
U3 = 0 U1 = -195
U2 = -155
V1 = 245 v2 = 200
V3 = 50
Cj
X 30 220
X X 170
250 X x
U1+ Vj
95
X 5 -145
X X -105
245 X x
Dij=Ui+Vj
X 25 365
X X 225
X X x
Optimal solution is
X-A 1 unit
Z – B- 2 units
Y to A – 3 units
Z – C (2 units)
Activity Questions
W1 W2 W3 W4 Capacity
F1 19 30 50 10 7
F2 70 30 40 60 9
F3 40 8 70 20 18
Req. 5 8 7 14
Problem 12
A B C D Availability
1 40 25 22 33 100
2 44 35 30 30 30
3 35 35 28 30 70
Req. 40 20 60 30
96
(Max profit = 5130 Rs)
13.Acompany has 3 plants p1, p2 p3 each producing 50, 100, 150 units of a similar
product. There are 5 ware houses w1, w2, w3, w4& w5 having demand of 100, 70, 50, 40
& 40 units respectively. Cost of selling one unit from various plants to ware houses differ
as given by cost matrix below
W1 W2 W3 W4 W5 ai
P1 20 28 32 55 70 50
P2 48 36 40 44 25 100
P3 35 55 22 45 48 150
dj 100 70 50 40 40
(Ans = 9240)
Problem 14
A B C D Supply
1 11 20 7 8 50
2 21 16 20 12 40
3 8 12 8 9 70
Demand 30 12 8 9 70
(AnsRs. 1130)
97
Unit III
NETWORK MODELS
The work on a project cannot be started until all its immediate preceding activities that
involve planning, procuring processing the inputs etc are completed.
Eg. Excavation for foundation should be done only after the site preparation; white washing
should be done after plastering
• Construction of a house
• Commissioning of factory
• Construction of ship
• New product launching
• Research to develop a new technology
Project management is generally applied for constructing items of public utilities, large
industrial projects, organizing mega events etc. Project management is considered to be
an important area in production scheduling mainly because many of the industrial
activities can be viewed as project management problem.
The problem of networks to find a certain course of action which minimize time, cost or
distance in performances of various activities.
Network techniques
98
Network techniques is a major advancement in management science. This technique is
based on the basic characteristic of all projects.
Different network techniques are PERT, CPM etc. The two network technique help
manages to plan, schedule, monitor and control large and complex projects.
1. It helps the management in planning the completed projects controlling work & plan
and updation of plan
2. It helps the management in reaching the goal with minimum time and least cost and
also in forecasting probable project duration & associated cost
3. Network techniques have resulted in better management control, better utilization of
resource and better decision making .
4. Network techniques have resulted in saving of time or early completion of project.
Networking diagram
The project can be broken into a number of distinct and well defined jobs called activities.
The beginning or end of each activities constitute an event of project
A graph drawn connecting various activities and events of a project is network diagram.
Each event is represented on it by a circle and each activity by arrow.
The arrow denotes the sequence of activities.
1. Event oriented diagram : Emphasis given on the events of project also referred as
PERT. Events are first selected and events in such network falls in a logical sequence.
Eg.CPM
99
1. Dividing the project into distinct activities
2. Estimating the time requirement for each activity
3. Establishing the precedence relationship among activities
4. Construction of arrow diagram (network)
Scheduling phase determined the start and end time of each and every activities which
The control phase uses the arrow diagram and time chart for continuous monitoring and
progress reporting.
Here A is an activity A
Number 2 initial node 2 3
Number 3 is terminal node
Dummy activities : Certain activities which neither consumers time non resources but are
used simply to represent a connection or a link between the events are known as dummy
activities . It is shown in the network by a dotted line.
1. To maintain the uniqueness in the numbering system as every activity may have
distinct set of events which activity can be identified.
2. To maintain proper logic in the network
B D
A Dummy
1. The Starting event and ending event of an activity are called tail event and head event
respectively
2. The network should have a unique starting mode (tail event)
3. The network should have a unique completion node (Head event)
4. No activity should be represented by more than one arc in the network
100
5. No two activities should have the same starting node and same ending
ending node
6. Dummy activity is an imaginary activity indicating precedence relation shop only.
Duration of dummy activity is zero.
Problem 1 Draw a network diagram to the following activities
Activity Time duration
1-2 2
1-3 4
1-4 3
2-5 1
3-5 6
4-6 5
5-6 7
Network diagram
101
Construct the network diagram
The earliest occurrence time or earliest event time (TE) is the earliest at which an event can
occur. Earliest occurrence of event 2 is denoted as E2.
E2
Latest event time (TL) : The latest allowable occurrence time or latest time by which can
event must occur
occur to keep the project on schedule. Latest occurrence of an event 2 is
denoted by L2.
1. Earliest start time (EST) : it is the earliest time which an activity can commence.
EST of activity 2-3
2 isE3
2. Earliest finish time (EFT) Earliest finish time of a activity is defined as the earliest
time by which an activity can be finished.
EFT = EST + Activity duration
3. Latest finish time (LFT)
Latest finish time for an activity is the latest time by which an activity can be
finished without delaying the completion of project.
pro
LFT of an activity =Latest expected time of head event
For activity 2-3
2 , LFT = L2
4. Latest start time (LST)
102
Latest start time of an activity is the latest time by which an activity can be started
without delaying the completion of project
Slack is a term associated with events . It denotes is flexibility range within which an
event can occurs.
Float ; Float is associated with activity time Float denotes the range within which the
activity starts time or finish time may fluctuate without affecting the completion of the
project
Types of float
1. Total float
2. Free float
3. Independent float
4. Interferring float
Total float
1. Total float is the time spent by which starting or finishing an activity can be delayed
without delaying the completion of the project
2.Free float : Free float is that portion of positive total float that can be used by an
activity without delaying any succeeding activity . The concept of free float is based on
the probability that all activities start as early as possible.
3 Independent float
It is defined as the excess of minimum available time over the required activity duration
Independent float is the amount of an activity could be delayed if preceding activity finish
at their latest and the subsequent activity start at their earliest
103
4 Interfering float =Total float – Free of float
Uses of float
Critical path
Critical path is the path having longest duration . The activities on the critical path are
shown by dark of double line arrows.
Critical path in the network diagram is a path on which all the activities have zero float.
Critical activity : critical activity is the activity lying in the critical path.
Features of CPM
Problem 1
A project schedule has the following characteristics
105
E10 = Max [E8+5, E9+7] = max (23,
(23, 17) = 22
Computation of TL values
L10 = L10 = 22
L9 = L10 - Duration (9,10) = 22-77 = 15
L8 = L10 – Duration (2,8) = 22-5 = 17
L7 = L8 – Duration (7,8) = 17-2 = 15
L6 = L8 – 1 = 17-1 = 16
L5 = Min (L8-4,
(L8 L7 – 8) = Min (12,7) = 7
L4 = L9 – 5 = 15-5 = 10
L3 = Min (L4-1,
(L4 L5-6) 6) = Min (9,1) = 1
L2 = L2 -1 = 10-1 = 9
L1 = Min (L2 – 4, L3-1)
1) = Min (5,0)=0
Finish Float
Activity Finish
Activity Start EST EFT = Start LST LST –
time LFT
EST EST
1-2 4 0 4 5 9 5
1-3 1 0 1 0 1 0
2-4 1 4 5 9 10 5
3-4 1 1 2 9 10 8
3-5 6 1 7 1 7 0
4-9 5 5 10 10 15 5
5-6 4 7 11 12 16 5
5-7 8 7 15 7 15 0
6-8 1 11 12 16 17 5
7-8 2 15 17 15 17 0
8-10 5 17 22 17 22 0
9-10 7 10 17 15 22 5
(11,3) (3,5) (5,7) (8,100 are critical activities with float zero
106
Problem 2 SDE KU UNI April 2014
The utility data for a network is given below. Determine the total float, independent
Activity 0-1 1-2 1-3 2-4 2-5 3-4 3-6 4-7 5-7 6
Duration 2 8 10 6 3 3 7 5 2 8
Solution
NETWORK DIAGRAM
Float Independent
Activity Start Finish Start Finish
Activity LST - float
time EST EFT = LST LFT
EST
0-1 2 0 2 0 2 0 -
1-2 8 2 10 8 16 6 -
1-3 10 2 12 2 12 0 -
2-4 6 10 16 17 23 7 0
2-5 3 10 13 22 25 12 0
3-4 3 12 15 20 23 8 0
3-6 7 12 19 12 19 0 0
4-7 5 16 21 22 27 6 0
5-7 2 13 15 25 27 12 0
6-7 8 19 27 19 27 0 0
E0 = 0
E1 = E0 + 2 = 0+2 = 2
E2 = E1+ 8 = 2+8 = 10
E3 = E1 +10 = 2+10 = 12
E4 = max (E2+6, E3 + 3) = max (16, 15) = 16
E5 = E2+ 3 = 10+3 = 13
E6 = E3 + 7 = 12+ 7 = 19
E7 = max (13+2, 16+5, 19+8) = 27
L7 = 27
107
L6 = L7 – 8 = 27-8 = 19
L5 = L7 – 2 = 27-2 = 25
L4 = L7 – 5 = 27-5 = 22
L3 = Min (19-7,
(19 22-3) = 12
L2 = Min (25-3,
(25 22-6) = 16
L1 = Min (16-8,
(16 12-10) = 2
L0 = L1 – 2 = 0
Critical
tical activity = 0-1
0 -3-6-7
108
Solution
E2 = E1+ 15 = 0+15 = 15
E4 = E3 + 8 = 18+8 = 26
E7 = 40+14 = 54
L7 = 54
109
L6 = L7 – 14 = 54-14 = 40
L5 = L6 – 3 = 37
L4 = Min (40-4,
(40 37-1) = 26
L2 = Min (18-3,
(18 37-5) = 15
L1 = Min (18-15,
( 15-15) = 0
Complete the earliest start, earliest finish, latest start and latest finish of each activity
of the project given below
Activity 1-2 1-3 2-4 2-5 3-4 4-5
Duration days 8 6 10 2 5 3
110
Network diagram
E1 = 0
E2 = E1 + 8 = 0+8 = 8
E3 = E1 + 6 = 8+6 = 14
To find TL values
L5 = 22
L4 = L5 – 3 = 19
L3 = L4 -3 = 16
L1 = Min (E2-8,
(E2 E3 – 6) = 1
Finish EFT Start Finish Float LST
Activity Duration Start EST
= EST LST LFT – EST
1-2 8 0 8 1 9 1
1-3 6 0 6 10 16 10
2-4 10 18 9 9 19 1
2-5 2 8 10 20 22 12
3-4 5 14 19 14 19 0
3-5 3 19 22 19 22 0
111
Critical activity = activity having zero float = 3-4-5
3 = 5+3 = 8
Critical path is represented in
Question No 5
Problem No 6
KU SDE NOV 2011 A project has the following activities and duration
112
a. What is the minimum completion period of work
E1= 0
E2 = 0+1 = 1
E3 = E2 +2 = 3
E5 = Max (E3 + 4, E4 + 1) = 7
L6 = 15
L5 = L6 -8= 15-8 = 7
L1 = L2 -1 = 0
113
Critical Path Diagram
Critical Path-1-2-3-5-6
An architect has been awarded a contract to prepare plan for an urban renewal project. The
114
b. Determine the critical path and calculate the total float and free float of each activity
115
Total float = LST – EST or LFT – LST
Critical activities
Draw the network and determine the critical path for the given data
Find the total float, free float and independent float of each activity
116
Calculation
E1= 0
E2 = E1+ 6 = 0+6 =6
E3 = E1+ 5 = 0+5
E5 = max (E3 + 4, E4 + 6)
= Max (9,22) = 22
E6 = max (E4 + 2, E5 + 9)
L6 = 31
L5 = L4-9 = 31-9 = 22
L4 = (L5-6, L6-9)
+= Min (18,13)= 13
L2 = L4 – 10 = 16-10 = 6
117
Independent float = EST of subsequent activity –LFT of preceding activity – Duration
Activity Questions
PERT is a management technique proposed for those project or operations which are of non
repetitive nature for these projects in which precise time determination for various activities
cannot be made
Assumption in PERT
1. Activity duration are independent ie time required for one activity has no relation with
another activities
2. PERT is used in those places where a project cannot be easily defined in terms of time
and resources.
3. Time duration in each activity is not a constant and hence un certain
118
Time estimates in PERT
This is the shortest possible time in which an activity can be completed under ideal
conditions. This particular time estimate represents the time in which the activity or
job can be completed if everything goes well with no problem or adverse condition
It is the maximum time that would be required to complete the activity. The particular
time represents the time it takes to complete a particular activity if everything went wrong
abnormal situation exists.
3. Most likely time estimate (tm) It is the time which the activity will take most
frequently performed a number of times. This time estimate a reflects a situation
()*+,-*,.
where condition are normal expected time based on 3 estimates.
/
Average time Te =
1. Identify the events and activities and prepare a suitable network for the given problem
2. Events are numbered in ascending order from left to right
3. Obtain the various time estimate for each activity . They are most likely tm,
pessimistic tp and optimistic to estimate
,)*+,-*,.
/
Calculate expected time te =
The activity with zero float are critical activities . Determine critical path through the
critical activities. Calculate the TE & TL for each event
5. Find the total expected duration time te by adding the time estimates for various
activities on critical path
CPM & PERT are both network analysis techniques used in planning and control of
projects. Both the techniques make use of network diagram for management of projects.
119
Difference between PERT & CPM
CPM PERT
1. CPM uses only single time estimate PERT uses multitime estimate for activity
for activity and does not consider duration (consider 3 time estimate
uncertainty. It is a determines to a uncertainity in time duration stochastic
model model)
2. CPM is used for repetitive jobs PERT is used for non repetitive jobs
3. CPM makes use of cost duration PERT analysis does not consider cost
relationship to arrive optimum
schedule
4. CPM is an activity oriented PERT is an event oriented
5. CPM is used for construction and PERT is used for planning scheduling
business problem research programme
1. Find the variance of time estimate of all determines variance of each activity =
012 134
5
σ = variance of project duration – sum of variances of time estimates of all critical
activities.
2. Find the probability of finishing the project on some fixed target by using the table of
normal distribution
Value of z = Due date- Expected date of completion /standard deviation of
critical path σ
σ = 678- )9 :;< ; =>7 )9 =< , =;? ;=, : , >7
Problem 11
For a project given below,Find1. The expected time for each activity
2.TG&TL values for all events
3. EST, EFT, LST,LFT values for all activities 4. Critical path
The time estimates are
Task A B C D E F G H I J K
Least
4 5 8 2 4 7 8 4 3 5 6
time
Greatest
6 9 12 6 10 15 16 8 7 11 12
time
Most
likely 5 7 10 4 7 8 12 6 5 8 9
time
120
Solution
Most
Optimistic time Pessimistic time Expected time
Task probably
to tp Te=1/6(to+4tm+tp)
timetm
A 4 6 5 5
B 5 9 7 7
C 8 12 10 10
D 2 6 4 4
E 4 10 4 4
F 7 15 7 7
G 8 16 8 9
H 4 8 12 12
I 3 7 6 6
J 5 11 8 8
K 6 12 9 9
121
Task Te EST EFT LST= LFT – duration LFT Float LST - EST
A(1-2)
2) 5 0 5 0 5 0
B(1-3)
3) 7 0 7 8 15 8
C(2-3)
3) 10 5 15 5 15 0
D(1-4)
4) 4 0 4 9 13 9
E(3-5)
5) 7 15 22 15 22 0
F(4-5)
5) 9 4 13 13 22 0
G(4-6)
6) 12 4 6 22 29 9
H(5-7)
7) 6 22 28 22 28 13
I(5-6) 5 22 27 24 29 0
J(6-8) 8 27 35 29 37 2
K(7-8)
8) 9 28 37 28 37 0
Critical Activities are 1-2,2-3,3-5,5-7,7-8
1
Critical Path A-CC-E-H-K
1-2-3-5-7-8
Problem 12
The following table lists the jobs of a network along with three time estimates
122
22
1.Draw the project network
3.What is the approximate probability that jobs on critical path can be completed in
a. 41 days b. 35 days
4.What is the probability that the probability that the project will not be completed within
45 days.
E1= 0
E2 = 0+7 = 7
E3 = 7+14 = 21
E4 = 7+5 = 12
E6 = 0+6 = 6
E7 = 6+11 = 17
123
E8 = Max (32+4, 17+18) = 36
L8 = E8 = 36
L7 = 36-18 = 18
L6 = 18-11 = 7
L5 = 36-4 = 32
L4 = 32-7 = 25
L3 = 32- 11 = 21
Critical Path=1-2-3-5-8
124
Variance of critical path = 4+ 16+4+1 = 25
σ=5
Z = t-t0/σ = 41-36/5 When z is the area under standard normal curve is 0.34
b. .t = 35 , z = 35-36/5 = -0.2
3. Probability that the job will not be completed with 45 days = 1-probability job
completed with 45 days when t = 45
Z = 45-36/5 = 1.8
t-te/σ = z
125
4. Suggests area for increasing efficiency and reduction of cost.
126
127
Critical activity = 1-3-5
Optimum length of critical path = 12+16 = 28
Variance of critical path
Activity Variance
015 − 94
=6
6
1-3
0 A4
5
3-5 = 144/6 = 24
SD = √30 = 5.47
1 1
. E
2. Z = σ
= = 2/5.47 = 0.3656
Z <0.3656
= 0.8656
Question 13
A small project is composed of seven activities where estimates are given in the following
table.
Activity Name of activity optimistic Most likely pessimistic
1-2 A 9 9 36
1-3 B 9 18 27
1-4 C 18 18 45
2-5 D 9 9 9
3-5 E 18 45 72
4-6 F 18 45 63
5-6 G 27 45 81
a.Find the expected duration of each activity
b. What is the expected project duration
Solution
te = to + 4 tm + tp/6
128
Calculation te = to + 4 tm + tp/6
9 F 04 G 944 F 36
Activity Expected time te
6
9 F 04 G 1844 F 27
A 13.5
6
18 F 04 G 1844 F 45
B 18.0
6
9 F 04 G 944 F 9
C 22.5
6
18 F 04 G 4544 F 72
D 9.0
6
18 F 04 G 4544 F 63
E 45.0
6
27 F 04 G 4544 F 81
F 43.5
6
G 48.0
Hence to determine the expected project duration we should draw a network diagram for
project, determine the expected lengths of various paths in the network using expected
times
A – D- G B E–G
B-
C–F
C-F
F = 22.5 + 43.5 = 66 day
129
Critical path is B- E – F with expected length of 111 days
Activity Questions
Problem 14
In a PERT network the critical path comprise of 6 activities. Its estimated duration of
days is given below.
Activity optimistic Most likely permistic
1 6 12 24
2 16 18 20
3 12 16 20
4 4 10 16
5 6 18 24
6 6 8 10
If the projects scheduled for completion within 84 days what is the probability of
achieving schedule(Ans: 0.6554)
Problem 15
For a small project, the critical path comprises of 4 activities whose estimates duration in
days are given below.
Activity optimistic Most likely pessimistic
1-2 3 5 7
2-4 7 9 11
4-5 - 0 -
5-6 6 8 10
a.20 days(Ans.0.0405)
Crashing of projects
The normal time of a project duration is the to time duration of all critical activities
together (ie critical path duration). The reduction in the project duration can be achieved
by reducing the time required for selected activities, from their normal time
The maximum reduction in the time possible for any activities is limited to its crash time
(ie beyond this time, reduction is not possible)
The progress of reducing the total project duration by reducing activity timings is known
as crashing of project
Procedure of crashing
130
1. Find the critical path and identify the critical activities
Calculate the new direct cost by adding the cost of crashing to the normal cost
5. As the critical path duration is reduced by crashing as per step 3, other path become
critical as we get parallel critical paths
This means that project duration can be reduced by simultaneous crashing of activities in
the parallel critical paths
6. By crashing as per step 3 and 4, one reaches a point when further crashing is either
not possible or does not result in the reduction of crashing of project duration (This
can happen when activities which can be crashed lie in the non critical path)
Total cost is got by adding corresponding over head cost to the direct cost and the
duration cost is got by adding the crashing cost cumulatively to the normal cost.
Problem 16
The following table gives duration in days and cost in Rs. Of the activities for a project
Activity Normal time Crash time Normal cost Crash cost
1-2 4 3 600 800
1-3 2 2 400 400
1-4 5 4 750 900
2-3 7 5 400 600
2-5 7 6 800 1000
3-5 2 1 500 650
4-5 5 4 600 850
4050
131
Solution
c. Optimum duration is the duration for which the total cost is minimum
minimu
To find the optimum duration, we crash the activities calculate the cost slope for each
activity using
800 600
Activity Normal time Crash time Normal Crash Cost Cost slope
200
4 3
1-2 4 3 600 800
900 750
1-3 2 2 400 400 0
150
5 4
600 400
1-4 5 4 750 900
200
7 5
1000 800
2-3 7 5 400 600
200
7 6
650 300
2-5 7 6 800 1000
150
2 1
850 600
3-5 2 1 500 650
250
5 4
4-5 5 4 600 850
132
Available time for crashing
Cost slope least for activity 2 and 3 and can be crashed for 12 days
1-2-3-5 1-2-5
To reduce the project duration to one more day (to reduce the days of duration to 10) we
have two choice
Crash 2 & 5 & 3 & 5 by one day (extra cost being Rs. 150+200 = 350)
133
1-2-3-5
1-2-5
Problem 17
The following table gives the activities and other relevant data of a project
Activity Normal Crash Normal Crash
1-2 4 3 120 100
1-4 6 4 300 500
1-3 2 1 60 120
2-4 5 3 300 500
134
3-4 2 2 200 200
2-5 7 5 230 350
4-5 4 2 300 480
Normal cost of project = 120 + 300 + 60 + 300+ 200 +230 + 200 (13 x 100) = 2710
135
Lowest cost slope pertaining to activity 2-4 crash 2 days from 2-4. Extra cost is Rs. 200.
Hence project duration is 10 days
Now there are 3 critical paths. That is path having duration 10 days
They are 1-2-5
1-2-4-5
1-4-5
Therefore we have to crash simultaneously from all these three paths
IIIrd stage crashing
Activity in critical path Possible crash days Cost slope
1-2 0 60
2-5 2 60
1-4 2 100
4-5 2 140
2-4 0 100
Now we have only one choice. That is crash 2 days from 2-5 and 4-5 simultaneously
1-2-4-5
1-4-5
1-3-4-5
IV stage crashing
136
After crashing maximum possible number of days project duration in 8 days with a total
cost of RS. 2870. But the total cost is minimum is RS. 2670 which the project duration is
10 day
Activity Question 18
A project has five activities and its is requested to prepare least cost schedule for all
possible duration from normal time to crash time
Activity Preceeding activity Normal Crash Normal Crash
A - 4 3 260 320
B - 8 5 300 510
C A 5 3 170 270
D A 9 7 220 300
E B,C 5 3 200 60
Features of crashing
Resource allocation
A resource is a physical variable such as man, material , machine, space or money that is
required for completing various activities or jobs of a project.
The network analysis PERT/ CPM can be successfully carried out if the availability of
resources is liberal or unlimited. But usually all necessary resources will not be available
in unlimited quantities. Availability of the some if the resource such as manpower and
material may be restricted. Availability funds, capital investment and heavy equipment
are the most important resources that need be allocated carefully
For a given network, the requirements of various resources are determined using the early
start schedule of each activity. There may be activities which are to be performed
simultaneously and may require common resources. The requirement of resources to
execute these simultaneous activities may exceed the available resources. Hence the
requirement of a particular type of resources may not be uniform all along the project
duration. The planning should be done on such a manner that resource are utilized in a
137
manner that resources are utilized in a more or less uniform manner. Large fluctuations in
these demand may cause problem in the project execution
1. Resource smoothing
2. Resource leveling
If the constraint is the total project duration, then the resource allocation only smoothens
the demand on resources. Hence the period for maximum demand is located and the
activities according to their float are shifted so that there is balance between availability
and requirement of resources. Proper utilization of floats can smoothen the demand of
resources to the maximum allocation is called resources smoothening
There are various activities in a project demanding varying levels of resources should not
go beyond the prescribed level. The operation of resource allocation is called resource
leveling.
In resources smoothening methods, the total project duration is not changed. But some of the
activities stage times are shifted by these available floats so that a uniform demand for the
resources is generated. However the resources are considered to be unlimited.
In resource leveling, the activity start times are so rescheduled that the peak demand for a
particular resource are considered to be limited. In resheduling the activities, the floats are
used. But its does not give the desirable results , the total project duration may be changed.
138
Maximum duration = 22 days
Consolidation
Resource Histogram
139
Maximum number of men required = 8
Problem 20
Find the maximum man power required for the project with same duration after rescheduling
source allocation by resource smoothing technique
Solution
TE values for all events
E0 = 0 E1 = 0+2 =2, E2 = 2+3 = 5
E3 = 2+4 = 6
E4 = 5+2 = 7 E5 = 6+4 = 10 E6 = 6+3 = 9
E7 = Max (10+6, 7+ ) = 16
E8 = 9+5 = 14 E9 = max (16+4, 14+4) = 20
TL value
L9 = 20 L8 = 20-4 = 16 L7 = 20-4 = 16 L6 = 16-5 = 11
L5 = 16-6 = 10 L4 = 16-6 = 10 L3 = Min (11-3, 10-4) = 6
L2=10-2=8 L1=Min( 6-4,8-3) =2 L0= 0
140
141
142
143
144
University Exam previous Questions in Unit 3 (Module 3)
1. What are the different types of floats associated with an activity in CPM Model(5
Marks)
5 .Explain the terms Optimistic, Pessimistic and most likely time estimates(5 Marks)
8 .Explain the different types of floats associated with an activity in CPM Model (5
Marks)
145
Unit IV
Queuing Theory
A queue is formed when the demand for a service exceeds the capacity to provide that
service.
1. To achieve the economic behavior between waiting time costs and service costs
2. Queuing theory can be used to estimate the different characteristics of the waiting line
such as average arrival rate, average service rate, average waiting time and average
time spent on the system
3. Queuing theory helps in taking decision about the optimum capacity so that the cost
of investment is minimum keeping the extent of queue within tolerancelimits.
1. Scheduling of aircraft
2. Scheduling of works and jobs in Production control
3. Minimization of congestion due to traffic delay in Toll booths
4. Scheduling of components to assembling line
5. Scheduling and routing of salesman
Queuing theory plays a very important role in management ultimate good of queuing
theory is to achieve an economic balance between cost of service and cost associated with
waiting, based on probability theory it attempts to minimize the extent and duration of
investment costs.
Queuing theory is able to provide with the estimated average time and intervals under
sampling method and helps in taking decision about the optimum capacity. So that cost of
investment is minimum keeping the extent of queue within tolerance limits.
Queing system
1. Single queue – single service point – There is only one queue and one service
station. This system is also known as single channel facility. In this system, the
146
customer waits in the only queue that is available till the service point is ready to take
him for servicing
2. Single queue multiple service points
In this system, there is only one queue there are severalservice points. A customer waits
in single queue until one of the services station is ready take him for servicing.
In this system, there are several queues and also several service station also known as
multi channel facility
Terminologies
1. Customer – Customer are these who waiting in the queue or receiving service
customers may be people, machine, ships, letters etc.
2. Queue – A group of item such as people, machine etc waiting for service in a service
station
3. Server – A server is a person or machine by whose service is rendered
4. Service station – Point where service is provided
5. System – Queue + service
6. Queue length – No of customers waiting in the queue
7. Time spent by customer – Time spent for waiting in the queue plus service time
8. Queuing system – Compress of arrival of customers waiting in queue, picked up for
service being served and departure of customers.
9. Average length of queue – Number of customer in the queue per unit time
10. Waiting time : Time upto which customer has to wait in the queue before taken to
service
Customersbehavior in queue.
1. Jockeying – Moving from one queue to another queue, when there are several
queues in the hope of reducing waiting time
2. Reneging – Joining the queue and leaving it after ward.
3. Balking – Customer have priority over others. Customers are served before others
regarding of their order of arrival
1. Most of the queuing models are complex cannot be easily understood. There is
always an element of uncertainty
2. Queue discipline also impose some limitation
3. In multiple channel queuing several times to practice from one queue forms
arrival for another make analysis more complex.
147
Characteristics of queing model
1. Arrival distribution
It is the input process. It represents the in which no of customer arrive at the system
called random arrival.
Arrival rate is defined as the rate at customers arrive to be service or the number of
customer arriving per unit of time
4.Queue output
In a single channel facilities the output of the queue does not possess any problem
for the customer who leaves after getting service. When the system is multi stage
channel facilities, because break down of a service station, can have repurcuison
on the queries.
148
2. Queue discipline is first come first served
3. Queue has infinite capacity
4. Arrivals are random and follow a poison distribution
5. No simultaneous arrivals
6. Service times are random and follow exponential distribution
7. No simultaneous arrivals
8. Traffic intensity is less than 1
Traffic Intensity δ
Traffic intensityδ is defined as the rate of mean arrival λ to mean service rate µ
Characteristics of queing system has been universally standardized in the format a/b/c
Eg. Simple queue which has poisson arrival distribution, exponential service time
distribution and single server represented as M/M/1
µ0µ λ4
λ
Queue length = E(nq) =
µ0µ λ4
λ
3. Average waiting time in the queue of an arrival E(w) =
4. Average time an arrival spend in the time system
149
E(v) = µ λ
5. Probability that the no an the queue and being served is greater than k is
λ
P (n>k) =(µ )k+1
Problem1
.
µ= x 60 = 24 customer/ hour
λ
Pn>6 (µ)k+1 = ( ) 6+1= 0.279
3. Expected time spent by the customer in the system
= (µ λ
= = 0.25 hours
= 0.25 x 60 = 15 minutes
Problem 3 Dec 20108- 10 Marks
150
Arrivals in a queue are considered to be poisson with an average time of 6 minutes
between one arrival and the next, length of queue service is an exponential variable
with mean time of 2 minutes
a. What is the probability that a person average at the queue will have to wait
b. What is the average length of queue which forms from time to time?
Solution
5 5
= 6 ,λ = person / minute = x 60 person = 10 person/ hour
=µ = 2, µ = ( ) person/ minute = 60 = 30 \]^_`a/hour
λ
λ
a. Probability that a person arriving at the queen will have to wait pn>k = ψ
Solution
λ= x 60 = 6 per hour
µ = x 60 = 20 per hour
b0b λ4
λ
3. Average waiting time in the queue =
b0b λ4
λ
= 3 minutes = 0.05 hours µ = 20
0 λ4
λ
= 0.05 λ = 10 per hour
151
5
5
= = 3/7
Problem 5
5. In a bank every 15 minutes, one customer arrives for cashing the cheque. The
staff in only payment counter takes 10 minutes for serving the customer on an
average.
state suitable assumption and find
1. Average queue length
2. Increase in arrival rate inorder to justify a second counter ( when the waiting
time of a customer is atleast 15 minutes, the management will increase one
more counter)
Solution
b0b λ4
=5
λ
For second counter to install, required arrival rate is 3.6 per hour
Problem 6
UNI SDE Nov 2011(10 Marks) A single man post office has 5 consumer for which the
single postman perform services. Assume that the service item is exponential andarrival rate
is poisson, find
152
Solution
λ = 4 per hour
b λ
3. Average time an arrival spends in system = =5 = 0.5 hr
= 30 minutes
Problem 7
KU Feb 2016 (5 Marks)
In a public telephone booth, the arrival on the average 15 per hour. A call on average takes 3
minutes. If there is just one phone, find the expected number of callers in the booth at any
time
Solution
λ = 15 per hour
µ = x 60 = 20 per hour
= = = 4 callers
Problem 8
At a certain petrol pump customers arrive according to poisson distribution with an average
time of 10 minutes between arrival. The time intervals between services at the petrol pump
follow exponential distribution and the mean time taken to service a customer is 4 minutes.
calculate
153
By how much should the flow of customer be increased to justify the opening of a second
service point if the management is willing to open the same provided the customer has to
wait for 10 minute for the service?
Solution
b
= 4 µ = = 0.25 per minute = 0.25 x 60 = 15 customers
5
b0b λ4 0 54
λ
1. Average queue length = = = 0.267 customers
5
2. Average no.of customers in the queuing system = b =0 54
λ
λ
=0.667customers
b0b λ4
λ
3. Average waiting time of customer in the queue = =
16/15(15-6)=2.67 minutes
4. Average time spend by the customer in the sytem=b λ
=1/(15-16)=
6.67 minutes
Mean arrival rate λ to determine which will result in an average waiting time of 1/6
5
= b0b λ 4 0
λ λ
= λ
λ1 = 10.71
Hence the mean arrival rate of a customer should go from the present rate of 6 customers per
hour to 10.71 customer per hour inorder to justify the opening of a second service point
Activity Questions
Problem 9
People arrive at a theatre ticket booth in a poison distribution arrival rate of 25 per hour
average time is exponentially distributed with an average time of two minutes
Calculate
1. Mean number in the waiting line
2. Mean waiting time
3. Utilization factor
154
Ans :(4.12, 10, 83%)
Problem 10
A postal clerk can service a customer in 5 minutes, the service time being
exponentially distributed with an average of 10 minutes during the early morning
slack period and an average of 6 minutes during the afternoon peak period. Calculate
Problem 12
Car arrives at a toll rate on a frequency according to poisson distribution with mean
90 per hour. Average time for poising through gate is 35 seconds. Drivers complaint
of long waiting time. Authorities are willing to decrease the passing time through the
gate to 30 seconds by introducing new automatic device. This can be justified only
ifunder the old system, the number of waiting cars exceeds 5. In addition the
percentage of gates idle time under the new system should not exceed 10. Can the
new device be justified
.
Traffic intensity in new system = b=
λ
= 0.75
Since the % of idle time in new system exceeds 10%, the new device is not justified.
Multichannel system
Queuingcost: costs associated with waiting in queue are known as waiting time costs
or queuing cost
155
1. Cost of idle time due to waiting
2. Loss of production when the equipment has to wait sometime to be repaired
3. Loss of sales due to customers going elsewhere upon seeing large queues.
4. Service cost : cost associated with service are known as service costs.
Problem 13
= x 60 = 6 machines / hr
5
Mean service rate = = hr
Service cost
156
There is only one server in the system
Hence the service cost of system consisting of cost of service. That is the average
hourly cost of clerk is Rs. 6
Total hourly cost of quality control dept
Problem 14
In a large factory fitter draw parts from the store, which is at present staffed by one
store keeper .The factory manager is concerned about the time spent by the fitter
getting parts from the store. He wants to know if the employment of storesAssistant
would be worthwhile. On investigation it is found
SolutionWe have to calculate the average number of fitters in the system before and
after introduction of stores assistant and cost comparison
A
λ
A
= = 4 fitters
Fitter cost per hour = Rs. 5
Hourly cost of queuing = Rs. 5 x 4 =Rs. 20/-
Hourly service cost = Cost of store keeper per hour = Rs 4/
Total Cost =20+4=Rs 24
157
A
A
= = 2 fitters
Conclusion
There is a net savings from the new system which is Rs. (10-3.50) = Rs. 6.50
Question 15A certain type of machine breaks down at an rate of Rs. 5 per hour. The
break downs are in accordance with the poisson distribution .Cost of idle machines
hours comes to Rs. 15/- per hour. Two repairman x and have been interviewed. X
changes Rs. 12 per hour and services the machine at the rate of 7 per hour where
asYchargesRs. 15 per hour and services the machine at an average rate of 9 per hour
which repairman services should be hired. (Assume that the workshift is of 8 hours)
Solution
Here the queue consists of machines waiting for service. To determine which
repairmen’s services should be used, We should calculate the total cost of queing
when x is employed.
158
E.
λ
=µ = = 2.5machines per hour
λ
This means that 2.5 machines hours will be lost in an hour due to waiting
In a work shift of 8 hours, machine hours lost will be 2.5 x 8 = 20 machine hours
cost of idle machine hours – Rs. 15 per hour
Hence queing cost of one workshift = 20 x Rs. 15 = Rs. 300
Hiring charges of X = Rs. 12 per hour
Hence service cost of one workshift = Rs. 8 x Rs. 12 = Rs 96
Total cost of queing and service = 300+ 96 = Rs. 39
When Y is employed
Mean arrival rate λ = 5 per hour
Mean service rate µ = 9 per hour
Expected number of machines in the system
λ
E(n) = = = 1.25 per hour
µ λ
This means that 1.25 machine hours will be lost in an hour due to waiting
In a work shift of 8 hours, machine hour lost will be 1.25 x 8 = 10 machine hours
Cost of idle machine = Rs. 15 per hour
Hence queing cost of one workshift = 10 x Rs. 15 = Rs. 150
Hiring charges of y = Rs. 15 per hour
Hence service cost of one workshift = 8 x Rs. 15 = Rs.120
Total cost of queuing& service = 150+ 120 = Rs. 270
Since the total cost of queuing& service is less when repairman Y is employed it is
more economical to employ Y
Parallel multi channel system
Multi channelsystemarequeuing systems where there are more than one service points
and one or more queues. Service points may be series or parallel.
M/M/K : In this case there is only one waiting line and several service points. Arrival
follows Poisson distribution and service time follows exponential distribution
whereKendells Notation as M/M >K where k stands for parallel service points
No. of unitsor customer in the system is represented by n
If n< k, there is no queue because all the arrival are being serviced.
If n > k queue is formed.
Traffic intensity of multi channel system δ=λ/ku
159
This is the queue model with
1. Arrival and departure following poisson distribution service rate following
exponential distribution
2. Single channel with infinite capacity
3. Service discipline :First come first served. There are single channel problems.
They are simple queuing problem. This is also called Birth and death model
4. Problems in previous pages belonging to model 1
Model II (M/M/I) : α/FCFS
This is same as Model I except that mean arrival rate and service rate are not constant.
Both are dependent on n in the number of units in the system
Arrival rate = λn service rate = µn
Consider the cases (1) λn = !* and µn = µ
λ
λ
1. Average queue length E(lq) = ELs - µ
2. Average waiting time in queue E9wq) = λ (Elq)
δc
dLO13K#LR !
3. Probability for n units in the system δn = e-δ
4. Probability that the unit is idling δo = e-δ
When λn = λ and µn = µ the model coincides with model 1
Question No 18A shipping company has a simple unloading built with ships arriving
in a poisson fashion at an average rate of 3 per day. The unloading time distribution
for a ship with an unloading crews is found to be exponential with mean unloading
!
time days. The company has a large laboursupply without regular working hours
and to avoid long waiting lines the company has a policy of using as many unloaded
crews as there are ships waiting in line or being unloading Under these conditions,
find
a.The average number of unloading crews working at any time
b .probability that more than 4 crews will be needed
Solution
Mean arrival time λn = λ = 3 ships per day (constant)
160
Mean service rate is not constant. It depends on waiting line µn = = 2n
e
F F F k
S gh δ S gh δ S gh δj S gh δ
! ! ! !
1 -f
=1-] l1 + 1.5 + + + m
. 0 . 4 0 . 4 j 0 . 4
5
=0.019
Model III (M/M/I) = N /FCS
This is a queuing model with (a) Poisson service (b) single channel (c) finite capacity
(d) first come first served
In this system, λn = λ when n<N
λn = 0 when n ≥ N
Where N is a fixed quantity such that arrivals will not exceed N
λ
µn = µδ = µ
Pn= 0
Probability forn units in the system
δ P*
δ
)δn for 0 ≤ n ≤ N and δ ≠ 1
δ P*
δ
Po =
δ P*
δ
= [s + 2s2 + …… Nsn]
Problem no.13
In a railway marshaling yard goods train arrive at the rate of 30 teams per day.
Assuming that interval arrival time follows an exponential distribution and the service
time (time taken to hump atrain) distribution is also exponential with an average of 36
minutes. The line capacity of the yard is to admit 9 trains only (these 10 lines, one of
which one is ear marked for the shunting engine to reserve itself from thecrust of the
hump to the rear of train
Calculate the following on assumption that 30 trains on an average are received in the
yard
161
d. Average waiting time in the system
e. Average waiting time in the queue
Solution
N = 90
λ = 30 trains per day
5 5
µ= trains per machine = x 60 x 24 = 40 per day
λ
δ=µ= = 0.75
2 δn *
δ
E(LS) = (δ+ 2δ2 + … 9δn)
1.32 hours
Problem 18
A petrol station has a single pump and space for no more than 3 cars ( 2 waiting and
one being served) A car arriving when the space is filled to capacity goes elsewhere
for petrol. Car arrive according to Poisson distribution at a mean rate of one every 8
minutes. Their service time has an exponential distribution with a mean of 4
minutes.
The proprietor has the opportunity of renting an adjacent piece of land, which would
provide space for an additional car to wait ( he cannot build another pump). The rent
would be Rs. 0.50 and the station is open 10 hours every day. Would it be profitable
to rent the additional space?
λ = 1 per 8 minute = A x 60 = 7.5 per hours
µ = 1 per 4 minute = x 60 = 15 per hour
λ E.
δ=µ = = 0.5
np
δ
Pn = Pn
δ
162
Maximum number in system = 3
Ien≤ 3 or N = 3
Consider when N= 3
.
0 . 4 0 . 4j
δ
P3 =
δ
x δ2= =0.067
Consider when N = 4
.
δ 0 . 4 0 . 4
δ
P4 = = = 0.032
δ
But rent for additional space = Rs. 10 per week which is more than 9.17
163
Unit 4
Simulation
University previous Questions Part A& Part B)
Identify problem
Validate model
Advantages of simulation
1.Simulation techniques can be used to solve the problem where the values of variables
are not known or partly known
2. Simulation methods are easier than analytical model
3. From the simulation model, data for further analysis can be generated
4. Effect of using the simulation model can be studied without using it in real situation
5. Simulation models are flexible
Disadvantages of simulation
Analysing where to locate factors or plants inorder to be able to distribute goods and
service at lowest rate
Level of machinery maintenance to minimize the service and break down cost
Examining the ship arrivals and unloading in harbour point to avoid forming of queues
165
Testing a series of inventory order policies to find the least cost order point
Montecarlo simulation
Monte Carlo simulation is a simulation technique which is used to solve the problems
by the application of random number. They are selected in such a way that every number
has an equal chance of probabilityof selection.
The basics of Monte carlo techniques is the random sampling of variables possible value.
For this technique some random number are required which may be converted into
random variates whose behavior is known from past experience. Monte carlo technique is
concerned with experiments on random number and provide solution to complicated OR
problem
This method uses random number for originating some data by which a problem can be
solved. The random numbers are used in creating a new set of hypothetical data of a
problem from past experience. If no pattern can be assumed for the data, then randomness
can be assumed. When past information is not available, it can be obtained by conducting
a preliminary survey. The data collected are plotted on a graph from which a cumulative
probability function is obtained.
In Monte Carlo method, the sequence of s random number are selected from the random
number table. The random number obtained as taken a decimal numbers and also the
probability obtained at random from the parent population. These probabilities are
plotted on the cumulative frequency curve of the given data. The value of x correspond
to each probability given by the random number is the desired random sample values.
1. Construct the model which represents the relevant characteristics of the real world
system
2. Determine the probability distribution for the value of interest.
3. Determine the cumulative probability
166
4. Select the random numbers using random number table
5. From the sequence of random numbe,r determine the corresponding variables of
interest.
1. Queing problem
2. Inventory problem
3. Risk analysis concerning major capital investment
4. Flexible budgeting
Problem 1(SDE Feb 2015)
1. Suppose the demand for a particular item is normally distributed with a mean of 175
units and standard deviation of 25 units / day. Simulate the demand for the next 20
days.
Solution
Determine the cumulative probability using the normal distribution table. Assign random
numbers corresponding to probability
167
Simulating demand for 20 days using random number given below
Day 1 Random number Demand x
1 59 180
2 65 190
3 29 159
4 13 150
5 56 178
6 12 145
7 63 206
8 61 205
9 70 192
10 75 195
11 54 177
12 90 210
13 96 222
14 26 156
15 43 173
16 07 128
17 08 129
18 40 172
19 79 208
20 45 174
Problem 2
A tourist car operator finds that during the past 100 days, the demand for the car had been
carried as shown below.
Trips per
0 1 2 3 4 5
week
No. of day 8 12 15 30 20 15
168
Simulation for 10 weeks
Demand(No.of
Weeks Random number
Trips)
1 9 1
2 54 3
3 42 3
4 01 0
5 80 4
6 06 0
7 26 2
8 57 3
9 79 4
10 52 3
The simulated demand for the cars for the new 10 weeks = 2.3
Explanation
To find the demand ,locate interval in which the random number of a week. The interval
shows the demand
Activity Question3
Suppose that the sales for a particular per day is Poisson with mean 5, then general 20
days of sales by Monte carlo method
169
Unit V
Game theory
Game Theory is a theory of conflict and it is mathematical theory which deals with
competitive situations
It is a type of decision theory which is concerned with the decision making in situations
where two or more rational opponents are involved under conditions of competition and
conflicting interests
Definitions
1.Game
Game is an activity carried out by two or more persons having conflicting interest who
agreed to play abiding by certain predetermined rules and limitations
2. Competitive games
This involves two or more intelligent persons parties actively participate to dominate over
the components strategies and make attempts to win over.
Here each participant has a set of finite or infinite choice of actions available to him and
anyone of these choices can be used freely at a given time
3.Strategy :
Strategy of a player is determined rule by which a player decide his course of action
duringthe game. That is a strategy for a given player is a set of rules or programmes that
specify which of the available courses of section, he should select at each play
6.Mixed strategy : A player is said to adopt mixed strategy when he does not adopt a
single strategy all the time but would play different strategies each a certain time. The
player would choose a particular course of action on the basis of probability.
The outcome of a game in the form of gains or losses to the completing player for
choosing different courses of action is known as pay off. In the game, the gain and losses
resulting from different moves and counter moves when represented in the form of a
matrix is known as pay off matrix
170
B1 B2
2 −3
r2 l m
0 1
Here A is the maximizing player and B is the minimizing player. Each element in the
matrix is the gain for A when he chooses a course of action against which B chooses
another course of action
Here each rows of the pay offmatrixindicates gain to A for his particular strategy. A is
called maximizing player and B is minimizing player.
Value of game
Value of game is the maximum guaranteed gain to the maximizing player. (A) if both the
player use the best strategies. It is the expected pay off of a play when all the player of the
game follow there optimum strategies.
Game involving two players is called two persons game and n person is called n person
game
A zero sum game is one in which the sum of payments to all the competitors is zero fo
every possible outcome of the game. Here the sum of the points won equals the sum of
points lost.
A non zero sum game in which the sum of the pay offs from any play of the game may
be either positive or negative but not zero.
171
3.Finite and Infinite game
A game without any limit on the number of moves or play is referred as infinite game.
A game with only two players in which the gain of one player is equal to the loss of other
player is called a two person sum games represented by pay off matrix in rectangular
form
1. A player may have infinite number of strategies. But we assume that there are
onlyfinitenumber of strategies.
2. Game theory ignores the presence of risk and uncertainly
3. The assumption that the gain of one person is the loss of opponent need not be true in
all situations
4. The assumption that the two persons involved in the game have equal intelligence
may not be true always.
Maximinprinciple
Here the maximizing player (A) lists worst possible pay off of all his potential strategies
and chooses the strategies which corresponds to the best. This is called maximum
principle .
Minimax principle
In this case, the minimizing player (B) lists his maximum losses from each strategy and
selects the strategy which corresponds to the least. This is called Minimaxprinciple.
Saddle point
A saddle point of a pay off matrix is the position in pay off matrix where the maximin
coincides with minimax. Pay off at the saddle point is the value of game
Example
172
Player B
3 2 4
Player A s−2 1 −3t
0 −2 3
Maximin = Max of row min = 2
Minimax – Minimum of column max = 2
Maximin = Minimax = 2 (A1, B2)
Saddle point is (A1, B2)
Value of game = 2
Problem 1
Player B
−2 15 −2
Player A s−5 −6 −4t
−5 20 −8
Value of game is -2 (A1, B1) (A1, B3)
Best strategy for player A is I
Best strategy of player B is I or III
Problem 2
15 2 3
Player A s 6 5 7t
−7 4 0
Row min 2 , 5, -7
Column max = 15,5,17
Max min= Minimax = 5 at (A2 B2)
Saddle point at (A2, B2)
Value of game = 5
173
Problem 3
9 3 1 8 0
6 5 4 6 7
u v
2 4 3 3 8
5 6 2 2 1
A company management and labout union are negotiating a settlement. Each has the
strategies
1. Aggressive approach
2. Logical approach
3. Legalistic approach
4. Conciliatory approach
The cost to the company for every paise of strategy choices are given below.
40 30 24 70
50 20 16 20
u v
80 4 20 10
−10 8 22 0
Arrive at the decision what is the value of game
Maximin=Minimax = 24 at (A1, B3)
Saddle point (A1, B3)
Union strategy = Aggressive approach
Companysstrategy Legalistic approach
Value of game = 24
Solution of mixed strategy problem
When there is no saddle point for a game problem, the minimax = maximin principle
cannot be applied to solve that problem. The strategies thus used to solve the problem is
called mixed strategy
1. Probability method
174
2. Linear programming method
3. Graphical method
Probability Method
This method is applied when there is no saddle and the pay off matrix has two rows and
two columns only. The player may adopt mixed strategies with certain probabilities.
Here the problem is to determine the problem of different strategies of both players
expected value of the game
$ '
Consider the following pay off matrix
l m
w x
Let p be the probability of A using strategy A1, probability for A using strategy A2 is 1-p
Let q be the probability of B using the strategy .probability of B using strategy B2 is (1-q)
LX yO X y X O
Value of game V = 0L*X4 q = 0L*X4 p= 0L*X4
0y O4 0y*O4 0y*O4
Problem 5
3 5
l m
4 1
There is no saddle point for this point since maximum and minimax
A1 = p A2 = 1-p
B1 = q B2 = 1-q
X O
P = 0L*X4 =0
0y*O4 * 4 0 * 4
=3/5
1-p =
X y
q = 0L*X4 =0
0y*O4 * 4 0 * 4
=4/5
1-q =
LX yO E
Expected value of game = 0L*X4 0y*O4
= =
z |
{ {
Strategy of A = (p, 1-p) =( , )
+
{ {
Strategy of B = (q, 1-q) = ( , )
175
}
{
Value of game =
Solve the following pay off matrix. Determine the optimal strategies and value of
game
5 1
l m
3 4
There is no saddle point since minimax and maximin are not equal
A1 = p A2 = 1-p
B1 = q B2 = 1-q
X O
p = 0L*X4 0y*O4
=0 * 4 0 * 4
= =
1-p – 1- =
=4-1/5=3/5
X y
q= 0L*X4 0y*O4
1-q=1-3/5=2/5
~$•€] `• ‚$ƒ] = 0
E
* 4 0 * 4
=
Problem 7
For the given pay off matrix, find the saddle point if it exists or not. Find the mixed
strategy for the player.
Player B
3 2
Player A l m
−1 4
X O
P = 0L*X4 0y*O4
a=3, b = 2 , c= -1, d= 4
*
0 * 4 0 * 4 5
= 1-p =1-5/6 = 1/6
176
X*y
q = 0L*X4 0y*O4
= 1-q =
X O *
Value of game = 0L*X4
0y*X4
= =
Solve the following pay off matrix. Determine optimum strategiesand value of game
Player B
2 8
Player A l m
10 6
Principle of dominance states that if the strategy of a player dominates over another
strategy in all conditions, hence the latest strategy can be ignored because it will not
affect the solution in any way. This is applied in the case of pure strategy and mixed
strategy
Dominance principle is useful in reducing the size of given pay off matrix.
Rule of Dominance
1.If all the elements in the row of a pay off matrix are less than or equal to the
corresponding elements of another row, then the latter dominates and the former is
ignored. (Remove the row having less values)
2. If all the elements in a column of a pay off matrix is greater than or equal to the
corresponding elements of a column, then the latter dominates and so former is ignored
(Remove the column having max values)
3. If the linear contributing of two or more rows (columns ) dominate a row (or
column),then the latter is ignored
case I : If all the elements of a row are less than or equal to the average of the
corresponding elements of two other rows , then the former is ignored
Case II : If all the elements of a column are greater than or equal to the average of
corresponding elements of two or other columns than the former is ignored.
By using dominance property we should try to reduce the size of pay off matrix to 2 x2
matrix
Using dominance property obtain the optimum strategies for both the player determining
the value of game
177
Problem 9
Player B
1 2 4 3 4
2 5 6 7 8
Player A u v
3 6 7 8 7
4 4 2 4 3
All the elements of column IV are greater than or equal to the corresponding elements
column I so we can elements column IV. The resultant matrix
Player B
2 4 3 4
5 6 3 8
Player A u v
6 7 8 7
4 2 8 3
All the elements of row 4 are less than the corresponding elements of row 3. So IVth row
can be deleted so we have
2 4 3 4
s5 6 3 8 t
6 7 9 7
All the elements in column V are greater than the corresponding elements of Column
1,So delete column V
2 4 3
s5 6 3 t
6 7 9
All the elements of row I and II are less than the corresponding elements of row 3, so
delete both rows 1 and 2
Resultant matrix is
3 „6 7 9…
Here the column I dominates II and III. So deleting column II and III
Value of game = 6
178
Solve the game whose pay off matrix is given
1 7 2
6 2 7
6 1 6
Each element is in row 3 is less than or equal to the corresponding elements in row 2.
Hence row 3 is deleted. Reduced matrix is
1 7 2
l m
6 2 7
Every element of column III is more than the corresponding elements of Ist column.
Hence column III is ignored.
1 7
l m
6 2
a= 1, b= 7, c= 6, d= 2
0LX yO4 E 5
V = 0L*X4 0y*O4
=0 * 4 0E*54
= =4
X O 5
P = 0L*X4 0y*O4
= = =
1-p = 1- =
0X y4 E
Q = 0L*X4
0y*O4
= = =
1 -Q =
| z
{ {
Optimum strategy for A is ( , ,)
Value of game = 4
Reduce the following game by dominance and find the value of game
3 2 4 0
3 4 2 4
4 2 4 0
0 4 0 8
179
Each element of row I is less than or equal to elements in row III. So row I is deleted.
3 4 2 4
Reduced matrix s4 2 4 0t
0 4 0 8
Each element of row I is less than or equal to elements in row III. So row I is deleted.
3 4 2 4
4 2 4 0
u v
0 4 0 8
Each element of Column I is greater than the elements in Coloumn s III. Hence So column I
is deleted.
4 2 4
2 4 0
u v
4 0 8
Column 1 is greater than the average of column 2&coloumn 3.Hence Column 2&3 dominates
and Column 1 is deleted
Row I is less than or equal to the average row of II and row III
4 0
Hence reduced matrix is l m
0 8
X O A
P= 0L*X4 0y*O4
= = 1-p = 1- =
X y A A
Q = 0L*X4 =0
0y*O4 *A4 0 * 4
= =
1-q = 1- =
;† ‡= + ˆ ‰ Š z| ‰
Value of game = 0;*†4
0‡*=4 | | z
= = =
|
Strategy of A is (z, z)
|
z z
Strategy of B is ( , )
Problem 12
180
8 10 9 14
A s10 11 8 12t
13 12 14 13
Resultant matrix
8 10 9
s10 11 8t
13 12 14
Reduced Matrix
10 11 18
l m
13 12 14
A3 „13 12 14…
OR
11 8 „
l m 12 14…
12 14
Value of game = 12
Activity Question 13
1 7 2 −1 −2 8
2 1 0 −2
s6 2 7 t s 7 5 −1t l m
1 0 3 2
5 1 5 6 0 12
Graphical method can be applied only if the pay off matrix is of the order 2 x n and m x 2
181
Hence A has only two strategies and B has n strategies. Let the strategies of A be A1 & A2.
Let p and 1-p be the probabilities with which player A uses his pure strategies. Player A will
select the value of p which to maximize the minimum excepted pay off.
For this we draw lines representing the strategies of B, say B1, B2 …. Bn. The lower
boundary of these lines will give the minimum expected pay off and the highest point in this
lower boundary will give the maximum expected pay off of player A and hence the optimum
value of P. Thus we get 2 x2 matrix. From the matrix, we find p & 1-p and expected value of
game.
Here B has only two strategies say B1 & B2. A has m strategies say A1, A2 …. Am. We
draw two vertical lines B1 & B2. Draw the lines A1, A2 …..Am by joining the respective
joints on B1&B2. The upper boundary of these lines gives the maximum of the loss and the
lowest point of this upper boundary is minimax point. The lines passing through the minimax
point represents the choice of A. This we get a 2 x 2 matrix. Solving this matrix, we get
strategies
A&B
Problem 14.
Player B
2 −4 6 −3 5
l m
−3 4 −4 1 0
Player A
2 4 6 −3
l m
−3 4 −4 1
Player A
182
Lower area bounded by lines B1, B2, B3 , B4 is LMNOP
N is the inter section of B1 and B4. So ignoring B2 and B3.Thus ignoring B2 and B3. Thus
the reduced matrix is obtained by avoiding B2 and B3.
2 3
l m
3 1
P=0
* 4 0 4
= 1-q =
Q=0 * 4 0 4
= 1-q =
LX yO E
V = 0L*X4 0y*O44
= 0 54
=
Solution
+ {
‹ ‹
As strategy = ( , )
+ {
Bs strategy are (‹, 0, 0, ‹ )
}
‹
Value of game =
183
Problem 5
Player Y
6 3 1 0 3
Player X l m
3 2 4 2 1
Lower area bounded by lines B1, B2, B3, B4, B5 is LMNOP. The maximum point of the area
is N. N is the point of intersection of B3 & B5. So ignoring B1 B2, B4 reduced
reduced matrix.
1 3
A1 = l m
4 1
X O
X2 optimal strategy p = 0L*X4 =0 4 0
0y*O4 4
=
X y
Q = 0L*X4
0y*O4 4 0 4
= = = 1-q =
LX yO
Value of game = 0L*X4 0y*O4
= =
184
z |
Solution to game is strategy Xs ({, {)
| z
{ {
StrategyYs (0,0, , 0, )
{
Value of game =
Problem 16
Solve graphically the game, whose pay off matrix of given below.
Player B
6 7
Ž 4 5‘
• •
Player A • 1 2•
• 2 5•
Œ 7 6•
Draw two vertical lines B2 and B1. Mark the values B1 & B2. Then draw A1, A2, A2, A4 ,
A5
Upper boundary LMNO. Lowest point on the upper boundary is N. N is the point of
intersection of lines A4 & A5. The game reduces to
2 5
l m
7 6
185
X O 5 E
P= 0L*X4 0y*O4
= = 5
1-p = 5
X y 5
Q= 0L*X4
0y*O4
= = 11/16 1-q = 5/16
LX yO
V = 0L*X4 0y*O4
= = 47/16
Solution
z
/
Strategy of player A (0,0,0, , 3/16 )
Strategy of B is (11/16,5/16 )
Player B
2 5
Ž−5 3 ‘
• •
• 0 −2•
•−3 0 •
Œ 1 −4•
+
Value of game =
{ }
As strategy ( |, 0 , 0, 0 |
)
z
Bs strategy (+, +)
186
187
Problem 18(UNI Feb 2015-10
2015 marks)
Use the notion of dominance to simplify the rectangle game with the following pay off. Solve
it graphically
Player K
I IIIII IV
Player
1 L
2
3
All the elements in column I is greater than column. Hence delete column I
4 6 4
4
s2 13 7t
5 17 3
Player K
II B1 IV B2
Player L
188
1
2
3
Solve the problem by graphical method
4
Strategy of L is (0, , 0, , )
Strategy of k is ( , , )
Reduced matrix is
2 7
l m
6 2
a= 2, b= 7, c= 6, d=2
LX yO E 5 A A
V= 0L*X4 =0
0y*O4 * 4 0E*54
= = =
X O 5
P = 0L*X4 y*O4
= 0 * 4 0E*54
= =
1-P=5/9
X y E
Q = 0L*X4 0y*O4
=0 * 4 0E*54
=
1-Q=4/9
189
190
1 +1 +1 0 0 0
B CB YB Y1 Y2 Y3 S1 S2 S3 Ratio
S1 0 1 1 -1 3 1 0 0 -1
S2 0 1 3 5 -3 0 1 0 1/5
∆j
S3 0 1 6 2 -2 0 0 1 1/2
-1 -1 -1 0 0 0
1 1 1 0 0 0
B CB YB Y1 Y2 Y3 S1 S2 S3 Ratio
S1 0 6/5 8/5 0 12/5 1 1/5 0 ½
Y2 1 4/5 3/5 1 -3/5 0 1/5 0 -1/3
2 8 1
S3 0 3/5 24/5 0 4/5 0 -2/5 1 3/4
5 5 5
0 0
191
1 1 1 0 0 0
B CB YB Y1 Y2 Y3 S1 S2 S3
Y3 1 ½ 2/3 0 1 5/12 1/12 0
Y2 1 ½ 1 1 0 ¼ ¼ 0
2 2 1
S3 0 1/5 16/15 0 0 -1/3 -7/15 1
∆j
3 3 3
0 0 0
Z = (1 x ) + (1 x ) + (1 x0) = 1
/ * /
Value of game = =1
Strategy of player B
Q1 = y1 V = x 1 =
Q2 = y2 V = x1=½
Q3 = y3V = 0 x 1 = 0
P1= x1v = x 1 =
P2 = x2V = x 1 =
P3 = x3v = 0
|
Solution to game = Asstrategy : (z, z, 0)
Bs strategy (|, |)
Value of game = 1
−5 2 0 7
As 5 6 4 8t
4 0 2 _3
192
Give the values and strategies adapted by A&B
Solution
Column Max=4
Row Min= 4
Saddle point= 4
Problem 24
1 7 2
s6 2 7 t
5 2 6
All the elements of row 3 is less than or equal to the corresponding elements of 2ndrow.Hence
3rd row can be deleted
1 7 2
6 2 7
All the elements of column 3 is greater than or equal to the corresponding elements of
istcolumn .Hence 3rdcolumn can be deleted
1 7
Hence the reduced matrix= 6 2
† = | /
Optimal strategy p = 0;*†4 0‡*=4
=0
*|4 0}*/ 4
= 4/10 1-p=6/10
193
† ‡ | }
Q = 0;*†4 0‡*=4
=0 *|4 0}*/ 4
= 5/10 1-q = 5/10
;† ‡=
Value of game = 0;*†4
0‡*=4
= 2- 42/-10 = 4
1. If saddle point exists, find the strategies of A and B are those corresponding to the
saddle point
2. When the saddle point does not exist, apply dominance rule and find the reduced
matrix
3. When the reduced matrix
a. 2 x 2 matrix – Apply probability method
b. 2 x m or n x 2 matrix – Apply Graphical method
c. mx n matrix- Apply LP techniques
194
Module V
Decision Theory
1. What is the significance of opportunity loss in decision making under risk.(Feb 2016-
5 Marks
2. Distinguish decision under uncertainty and decision under risk Feb 2015-5 Marks
3. Write short Notes on Opportunity loss-June 2014 -5 Marks
4. Explain decision making under risk and decision making under uncertainty -5 marks
Nov 2013
5. Explain the concept decision making under risk-Nov 2012-5 Marks
6. Discussion in details the different decision environment-Nov 2011 –10 Marks
7. Write short notes on a. Decision Tree Approach b .Laplace Transform
Decision theory
Decision theory is a term used to apply to those methods for solving decision problems in
which uncertainty plays a critical role
1. Decision maker
2. Number of alternatives
3. Different states of nature
4. Outcome of decision
Decision maker :Decision makers is charged with responsibility of making the decision
Course of Action (acts):Acts are alternative course ofactions or strategies that are
available to the decision maker. The decision involves selection among two or more
alternative courses of action. The problem is to choose the best of these alternatives to
achieve an objective.
Events are occurrences which affect the achievement of the objective. They are also
called the states of nature. The events constitute a mutually exclusive and exhaustive set
of outcomes which describe the possible behavior of the outcome in which decision is
made.
Outcome : When the decision makes selects a particular act under a particular state of
nature, the result obtained is called outcome.
195
Pay off : pay off can be interpreted as the outcome in quantitative form when the decision
make adopts a particular strategy under a gain or loss of each such outcome. The pay off
can also be based on cost or time.
An opportunity loss incurred because of the failure to take the best possible decision
,Opportunity losses are calculated separately for each state of nature that might occurs.
Given the occurrence of a specific state of nature. We can determine the possible act.
Opportunity loss = pay off that Act – pay off for the best act that have been selected.
In this case, decision makes know with certainity the consequences of every alternatives
or decision choice. The decisionmaker presumes that only one state of nature is relevant
for the purpose. He identifies this state of nature takes it for granted and presumes
complete knowledge as to its occurance
196
When the decision makes faces multiple states of nature, but he has no means to arrive at
probability values to the likelihood or chance of occurrence of these states of nature, the
problem is a decision problem under uncertainity
Eg., When a new product is introduced in a market. These decisions under uncertainty are
taken with even less information than decisions under risk. The following choices are
available before the decision makes in substitution of uncertainty.
A1 A2 A3
• • •
Maximum pay off under each decision A1, A2,A3 are respectively
respectively
In this criteria, the decision makes ensures that he would earn no less (or pay no
more) than some specified amount
197
In this the decision makes lists down the minimum outcome within every course of
action or alternative strategy and then selects the strategy with the maximum number
a. Locate the minimum pay off value in the cost or loss ( or maximization in the case
of profit) corresponding to each alternative strategy
b. Select an alternative with the best anticipated pay off value (minimum for profit
and maximum for cost or loss)
In this criteria ,the decision maker is conservative about the future and always
anticipates worst possibleout come (Minimum for profit and maximum for cost or
loss) it is called a pessimistic decision criterion. This criterion is also known as
waldian criterion
State of
A1 A2 A3
nature
S1 -80 -60 -20
S2 -30 -10 -2
S3 30 15 7
S4 75 80 25
• • •
A 5
Minimum loss incurred
Hence the decision makes should take a decision for A3. A 3 is the alternative to be
selected.
Since the probabilities of states of nature are not known , it is assumed that all states
of nature will occur with equal probability ie each state of nature has assigned an
equal probability
The expected pay off is worked out on the basis of probabilities. The act having
maximum expected pay off is selected. Each state of nature has assumed a
probability of 1/n of occurrence where is the no of states of nature.
This criteria is also known as criteria of in sufficient reason , because except in few
cases some information oflikehood of the occurrence of the state of nature is
available.
20 12 25
Events E1s25 15 30t
30 20 22
We associate equal probability for each event is 1/3 so the expected pay off are the
average of pay offs under each set
198
A1 = [20+25+30] = 25
A2 = [12+15+20] = 15.67
A3 = [25+30+22] = 25.67
This criterion suggests that a rational decision maker should be neither completely
optimistic nor permistic and therefore must display a mixture of both.
Di= M i α + (1-α) m1, where Mi is the maximum pay off ith strategy and mi is the
minimum pay offofith strategy
Decision maker will specify the value of α depends upon the level of optimism
Example
Acts
20 12 25
Events s25 15 30t
30 20 22
Assume α = 0.6
For A1 max pay off = 30Minimum pay off = 20
D1 = (0.6 x 30) + (1-0.6) 20 = 26
A2 max pay off = 20 minimum pay off = 12
D2 = (0.6 x 20) + (1-0.6) 12 = 16.8
A3 max pay off = 30, minimum pay off = 22
D3 = (0.6 x 30) + (1-0.6) = 26.8
Since D3 is more, use select the act A3
199
5. Criteria regret or minimax criteria
Regret table
Minimax is just opposite to maximax. Application of the minimax decisions requires
a table of losses instead of gains. These losses are the costs to be increased or
damages to be suffered for each of the alternative act and states of nature .The
minimaxrule minimizes the max possible loss for each course of action (minimax
observation of minimum of maxima)
Under each of the various act, there is a maximum loss and the act that is associated
with the minimum of various maximum loss is the act to be undertaken according to
0 4 10
minimax criteria
Example s 3 0 6t
18 14 1
• • •
A
A3 is the choice selected :
Solution
200
Maximum of maximax = 250 which relates to strategy I
Minimax criteria
Regret tableTable 2
Level of sales I II III
20000 250-250 250-40 250-60
10000 25-15 25-20 25-25
2000 10-10 10-5 10-3
Regret tableTable 3
Level of sales I II III
20000 0 210 190
10000 10 5 0
2000 0 5 7
6. Maximin(from table I)
7. Laplace criteria
201
Given α = 0.6 (1-α) = 0.4
Problem 2
The marketing dept of the company worked out the pay off in terms of yearly new
profits for each of the strategies on these events. This is represented on the following
table .
Strategies State pay off
N1 N2 N3
S1 700 300 150
S2 500 450 0
S3 300 300 300
Which strategy should be executive concerned choose on the basis of
a. Maximin extra (b) maximax criteria
c. Minimax regret criteria d. Laplace criteria
Solution
Rewriting the pay off table (with strategies in columns)
State of nature S1 S2 S3
N1 700 500 300
N2 300 450 300
202
N3 150 0 300
Maximin criteria
Minimum pay for S1, S2, S3
S1/150 , S2/0 , S3/300
Maxima of the minima = 300 which relate to S3.
Hence the executive should choose option S3
B, Maximax criteria
Maximum pay off for S1, S2, S3 are
S1/ 700, S2/500, S3/300
Maximum of the maxima = 700 which relate to S1
Hence the executive should choose option S1
c.Minimax regret criteria
Opportunity loss table
N1 S1 S2 S3
N2 700-700=0 700-500=200 700-300=400
N2 450-300=150 450-450=0 450-300=150
N3 300-150=150 300-0=300 300-300=0
S1/150, S2/300,S3/400
Hence the executive should choose strategy S1, since it minimizes the maximum
strategy
d. Laplace transform
Since the average is highest for strategy1 , the executive may select strategy S1
203
Problem 3
A company has an opportunity to computerize with department. However existing
personal has job security union agreement. The cost of three alternative
programmesfor the change over depends upon the attitude of the union are estimated
below.
Solution
General training Selective training Hiring new employee
Antagonist 940-940=0 940-920=20 940-900=40
Passive 820-810=10 820-800=20 820-820=0
Enthusiastic 860-700=160 860-710=150 860-860=0
Max opportunity loss is S1/160, S2/150, S3/ 40
Minimax of the max opportunity loss is relating withS3 = 40
Hence hiring new employee is the strategy followed since it minimise the
maximum strategy
The objective is to optimize the expected profit or minimize the opportunity loss
The following were the criterion followed for decision making under risk
Expected monetary value (EMV) for a given course of action is the weighted average pay
off which is the sum of pay off for each course of action multiplied by the probabilities
associated with each state of nature.
204
Pi – probability of occurrence state of nature
Pj – pay off associated with the state of nature Ni and course of action Sj
The decision makes then selects from the available alternatives, the act that lead to
optimum expected outcome (act with optimum EMV)
Optimal means maximization in the case of profit or income and minimization in the case
of cost, expenditure etc
30 20
l m
35 30
Let the probabilities for the states of nature S1 & S2 respectively 0.6 and 0.4
The difference between the greater pay off and actual pay off is known as opportunity
loss .Under the criteria, the strategy which has the minimum expected opportunity loss
(EOL) is chosen.
Expected opportunity loss (EOL) is defined as the difference between the highest profit
(or pay off) for a state of nature and actual profit obtained for a particular course of action
taken. EOL is also called expected value of regret.
EOL is the amount of pay off that is lost by not selecting the course of action that has the
greatest payoff for the state of nature that actually occur.
When the probabilities for various states of nature are known, it is possible to calculate
the expected for each course of action
EOL = (x1+p1) + (x2 x p2) + …. Where x1, x2 are losses and p1, P2 …. Are the
probabilities of State of nature. These expected losses are known as expected
opportunity loss (EOL)
205
Consider the following example. Given below is an opportunity loss table. A1 and A2 are
strategies S1 and S2 are states of nature.
A1` A2
S1 0 10
S2 2 -5
Let the probabilities for the state be 0.6 and 0.4
EOL of A1 = (0 x 0.6) + (2x 0.4) = 0.8
EOL of A2 = (100 x 0.6)+ (-5 x 0.4) = 4
EOL for A1 is least. Therefore the strategy A1 may be chosen
Problem 4
ABCD has invented a picture cell phone. It is faced with selecting one alternative out of
the following strategies. Manufacture the cell phone take royalty from another
manufacturer, sell rights for the invention and take a lump sum amount. The probabilities
of demand being high, medium and low are 0.25, 0.44 are respectively. The profit table is
given below
Demand Profitability Manufacturer Royalty Sell rights
High 0.25 20 60 50
Medium 0.40 50 40 50
Low 0.35 -10 20 50
Solution
We apply equally likely or laplace criteria for estimating alternatives
Expected pay off of different courses of action
Expected pay off for Manufacturer
E1 = (0.25 x 20 ) + (0.4x 0.50) + (0.35 x -10) = 21.5
Expected pay off for loyalty
(0.25 x 60) + (0.4 x 40) + (0.35 x 20) = 38
Expected pay off for sell rights
(0.25 x 50) + (0.4 x 50) + (0.35 x 50) = 50
Expected pay off is maximum in the case of alternatives sell rights. Hence we may
accept the alternative sell rights as an alternative the company select.
Problem 6
A management is faced ith the problem of choosing one of the products for
manufacturing. The profitability matrix after the market research for the two products as
follows.
Act State of nature
Good Fair Poor
Product A 0.75 0.15 0.10
206
The profit that the management can make for different levels of market acceptability of
the product are as follows
The profit that the management can make for different levels of market acceptability of
the product are as follows.
Good Fair Poor
Product A 35000 15000 5000
Product B 50000 20000 30000
Calculated the expected values of choices of alternative and advise the management
Solution
Product A product B
Pay off x profitability pay off x probability
Good 35000 x 0.75 = 26250 50000 x 0.6 = 30000
Fair – 15000 x 0.15 = 2250 20000 x 0.3 = 6000
Bad 5000 x 0.1 = 500 30000 x 0.10 = -300
Since the expected pay off (EMV) for product B is greater product B should be
preferred by management
Problem 7
Given below is the regret table
Act Events of state of matrix
A1 350 0 250
A2 0 300 0
A3 450 100 100
Suppose that the probability of events in these table are P(E1) = 0.3 P(E2) = 0.45 P(W3)
= 0.25
207
350 x 0.3 = 105 0 x 0.3 = 0.4 450 x 0.3 = 135
0 x 0.45 = 0 300 x 0.45 = 135 100 x 0.45 = 45
250 x 0.25 = 62.5 0 x 0.25 = 0 100 x 0.25 = 25
208
209
18. A grocery stores with bakerydept is faced with the problem of how many cakes to buy
inorder to meet days demand. The grocer prefers not to sell of goods in competition with
fresh products. Left over cakes are therefore a complete loss. On the other hand, if a
customer desires a cake and all of them have sold, the disappointed customer will be else
where and sales will be lost. The grocer has therefore collected information on past sales
or a selected 100day period as shown in the table below.
Sales per day No., of days probability
25 10 0.1
26 30 0.3
27 50 0.5
28 10 1
Total 100 1.0
Construct the pay of table. What is the original number of cakes that should be bought
each day. Find and interpret EVPI. A cake costs Rs. 0.80 and sells for Rs. 1
Solution
Let A1, A2, A3, A4 stands for strategies and S1, S2, S3, S4 stands for the states of nature
S1, S2, S3, S4 represents the demand for 25, 26,27, 28 values.
Conditional pay off values and pay value are obtained as shown below.
A2 S1 = 2.6 x 1 – 25 x 0.8 = 6
A3S1 = 27 x 1 – 2.5 x 0.8 = 7 A1S4 = 25 x 28 x 0.8 =2.6
A1S2 = 26 x 1 - 26 x 0.8 = 7 A4S2 = 28 x 1 – 2.6 x 0.8 = 5.2
A1S2 = 26 x 1 – 26 x 0.8 = 5.2 A2S4 = 26 x 1 – 26 x 0.8 = 5.2
A3S2 = 27 x 1- 26 x 0.8 = 6.2 A4S3 = 28 x 1 – 2 7 x 0.8 = 6
A1S3 = 25 x 1 – 2.7 x 0.8 = 3.4 A3S4 = 27 x 1- 26 x 0.8 = 4.6
A3S3 = 27 x 1 – 27 x 0.8 = 5.4 A4S4 = 28 x 1- 28 x 0.8 = 5.6
Maximum EMV is act A4 which is equal to 6.06. This according to EMV decision
criteria the store would stock 28 cakes
210
For calculating expected value with perfect information select highest pay off for each
state of nature(ie each row ) and calculate expected values
Pay off Probability Pay off x probability
8.6 0.1 0.86
6.2 0.3 1.86
6.4 0.5 3.20
5.6 0.10 0.50
6.36
Problem 11
Maximax ,minimax, Maximin , Laplace, EMV, EOI Given probabilities for various
events are 0.3 , 0 .4 &0.30 respectively
5 −1 −17 0 6 22
s18 28 30 t s10 0 2 t o.3
25 30 80 55 50 0
• • •
Maximax criteria
A
Maximum pay off =
Maximum of maxima = 80 (which relates to A3) Optimal Act is A3
• • •
Minimax criteria
From regret table max loss =
Minimum of the maxima = 22 which relates to optimal act is A3.
• • •
3.Maximin criteria
E
Minimum pay off from pay off table =
Maximum of these minima = 5 which relates toA1 .optimal act is A1.
Laplace criteria
211
Average pay off of the acts A1 = (5+ 18+ 25) = 16
A2 = (-1+28+30) = 19
A3 = (-17+30+80) = 31
Optimal act is A3
5.EMVcriterian
6. EOL Criterian
Problem 12
A small ink manufacturer produces a certain type ink at a total average cost of Rs. 3/ per
bottle and then sells at a price of Rs. 5 per bottle. The ink is produced over the week end
and is sold during the following week. According to the past experience the weekly
demand has never been less than 78 or greater 80 bottles in this place.
You are requested to formulate pay off
Solution
212
The different states of nature are selling 78 units, 79 units 80 units respectively are S1,
S2, S3. The different courses of action are selling 78 units 79 units and 80 units
represented as A1, A2, A3
Problem 13
A firm manufacturers three types of products .The fixed and variable costs are given
belowFixed cost(Rs) Variable cost (Rs)
Product A 25000 12
Product B 35000 9
Product C 53000 7
The likely demand (units of products) is given below
Poor demand = 3000
Moderate demand = 7000
High demand = 110000
If the sale price of each type of product is Rs. 25 .Prepare the pay off matrix
Solution
Let D1, D2, D3 be the poor, moderate and high demand respectively
Then pay off = sales revenue – cost
The calculation of pay off ( in thousand) for each pair of alternative demand (course of
action) and types of product (states of nature) are shown below.
D1A = 3 x 25 – 25 – 3 x 12 = 144
D1B = 3 x 25 – 35 – 3 x 9 =13
213
D1C = 3 x 25 – 53 – 3x 7 = 1
D2A = 7 x 25 – 25 – 7 x 12 = 66
D3C = 11 x 25 – 53 -11 x 7 = 145
Pay off values are shown in the table
Alternative demand
Product type A D1 D2 D3
A 14 66 118
B 13 77 141
C 1 13 145
Problem 14A news paper boy has the following probability of selling of magazine
10 0.10
11 0.15
12.1 0.20
13 0.25
14 0.30
Cost of copy is 30 paise and sale price is 50 paise. He cannot return unsold copies. How
Solution
We can apply either EMV criterion or EOL criterian. Let us apply EMV criterian for which
we have to calculate pay off number of copies ordered are different courses of action. The
copies
214
215
216
conditional profit table
State of nature :High demand and moderate demand and course of action expand or
modernize
State of nature Expand Moderate
High demand N1 12-8=4 6-5 =1
Moderate demand N2 7-8=-1 5-5=0
Cause of Conditional
State of nature Probability Weighted
action profit
High demand N1 0.35 S1 A 4x 0.35 = 1.40
0 x 0.65 = 0
Moderate demand N2 0.65 S2 0
EPPJ = 1.40
The optimal EMV is Rs. 0.75 million corresponding to course of action S1
EPVI = EPPI – EMV (S1)
= 1.40 – 0.75 = Rs. 0.65 million
The company could get a perfect information of demand it should consider paying
uptoRs. 0.65 million for an information
217
The decision maker seeks to minimize the expected opportunity loss he must select a
cause of action S1 s it produces the smallest expected opportunity loss
Activity Question 16
The manager of a flower shop promises its customers deliverywithin four houses on all
flower orders. All flowersare purchased on the previous day and delivered to Packer 8
Am the next morning. The daily demand for roses as follows.
Dozen of roses 7 8 9 10
Problem 17
The following is a pay off (in rupees) table for three strategies and two states of nature
Strategy (S1) N1 N2
S1 40 60
S2 10 -20
S3 -40 150
Select a strategy using each of the following decision criteria
a.Maximax
b.Minimax regret
c. maximin
d. Minimum risk assuming qui-probable states
For making use of Bayes principle in the statistical decision problem, the decision maker
has to assign probabilities to each state of nature. These probabilities represent the
strength of decision makes belief is a subjective evaluation regarding the likelihood of
occurrence of various states of nature.
After determining the probabilities, the Bayes principle must be used phase wise. The
three phases are
1. Prior analysis
2. Pre posterior analysis
3. Posterior analysis
218
A decision maker assigns probability to various events which is his subjective evaluation
of probability based on experience of past performance.
performance. When these prior probabilities
are used the procedure is known as prior analysis.
If prior analysis is reveals a high EVPI, additional information are to be obtained. Prior
probabilities maybe then, revised on the basis of these additional information
info .By
applyingbayes theorem of probability the revised probabilities arecomputed. These
probabilities are known as posterior probabilities .A further analysis of the problem
using these post probabilities give new expected pay off. The revised of the problem is
known as posterior analysis.
Preposterior analysis is done to assess the value of sample information as against the
expected of perfect information even before selecting a sample additional information.
This analysis involves of the probabilities
probabilities using Bayesr Theorem.posterior analysis
involves arriving at a decision after revision probabilities.
Decision Trees
The decision tree consists of nodes and branches. The nodes are two types, decision nodes
and chance node. Course of action or strategies originate from the decision nodes or main
branches. At the terminal of each main branch, there is a chance node. From these chance
nodes, chance events emanate in the form of subbranches
The respective pay offs and probabilities associatedwith alternative courses and chances
events areshown along with the subbranches.
At the terminal of the sub branches are shown the expected value of outcome
219
Here A1, A2, A3 are strategies E1, E2, E3 are the events and O11, O12, O22, O31 are the
outcomes.
A decision tree is highly useful to a decision maker in multi stage situation which involve
a series of the decision each dependent on the preceding one working backward from
future to present, we are able to eliminate unprofitable branches and determine optimum
decisions. The decision tree analysis allows one to understand, simply by inspection,
various assumptions and alternatives in a graphic form which is much more easier to
understand then the abstract analytical form.
The advantage of the decision tree structure is that complex managerial problems and
decisions of the chain like nature can be systematically and explicitly defined and
evaluated.
Problem 17
There are 40% chance that patient admitted to the hospital is suffering from cancer. A
doctor has to decide whether serious operations should performed or not. If the patient is
suffering from cancer and the serious operation is performed, the chance that he will
recover is 70% otherwise it is 35%. On the other hand, if the patient is not suffering from
cancer and serious operationis performed, the chance that he will recover is 20%
otherwise it is 100%. Assuming that recovering and death are the only possible
results.Construct an appropriate decisiontree . What decision should the doctor take.
Solution
Let P1, probability that patient shall recover from operation= 0.28 + 0.12 = 0.4
P2 = Probability that patient will recover with any operation= 0.14 + 0.60 = 0.74
Since P2 > p1 doctor should not take a decision for operation to the patient .
220
Problem 18
A firm owner is seriously considering of drilling a farm well. In the past, only 70% of the
wells drilled were successful at 200 feet of depth in the area. Moreover on finding no
water at 200 ft, some persons drilled it further upto 250 feet but only 20% struck water at
250 ft. The prevailing cost of drilling is Rs. 50 per foot. The farm owner hs estimated that
in case he does not get his own wells.He will have to pay Rs. 15000 over the next in pvt
term to buy water from the neighbor.The following decisions can be optimal
Draw an approach decision tree determine the farm owners strategy under EMV approach
221
At D2 point
Decision A. Drill upto 250 ft b. Do not drill
Event a. No water b. Water
Probabilities are 0.2 & 0.8
EMV for drill upto 250 feet
= (12500 x 0.2 ) + (27500 x 0.8) = 24,500
EMV for do not drill =25000 (from the tree)
EMV is smaller for the act drill upto 250 ft. So it is optimal act.
At D point
The decision are drill upto 200 feet and do not drill.Events are same as those of D2 point
probabilities are 0.7 and 0.3
222
223
224
At point D1
Decisions are (i) Development drug
(ii) Accept proposed it
Events are (i) successful (ii) Not successful
Probabilities 0.6, 0.4
Develop new drug
EMV = Expected PV = (23000 x 0.6) = (0 x 4) = 13800
Accept proposal H
EMV = Expected PV = 12000 x 1 = 12000(Given)
Using EMV criterion, optimal decision at D1 is to develop and market new drug
Assignment questions of OR
All questions are compulsory
1. Describe the various OR software used in an industry
2. Solve the problem by simplex method
Minimise Z = 20x1+ 24x2 + 18x3
Subject to constraints
2x1 + x2 + x3 ≥ 30
X1 + x2 + x3 ≥ 20
X1+ 2x2+ x3 ≥ 20
X1, x2, x3 ≥ 0
3. A company has 3 plants P1, P2, p3 each produces 50, 100, 150 units of a similar
product. There are 5 warehouses W1, W2, W3, W4 and W5 having demands of 100,
70, 50, 40and 40 units respectively. Cost of selling oneunitfrom various plants to
ware houses differ as given by cost matrix below.
4. The manager of a flowershop promises its customers delivery within four houses on
all flower orders. A flowers are purchased on the previous day and delivered to
pacrker by 8 am the next morning
225
Dozen of roses 7 8 9 10
A manager purchases roses for Rs,=10/- per dozen and sells them for Rs. 30. All unsold
roses are donated to a local hospital. How many dozen of roses should pacrker order each
evening to maximize the profit. What is the optimum expected purchase.
5. In a railway marshaling yard, goods train arrive at the rate of 30 trains per day.
Assuming the interval arrival time follows an exponential distribution and severe time
(time taken to hump a train) distribution is also exponential with an average of 36
minutes. The line capacity of the yard is to admit 9 trains only (There being 10 lines
one of the which one is ear marked for shunting engine to reverse itself from the crest
of the humps to the rear of train
Calculate the following on assumption that 30 trains on an average are recieved in the
yard
6. A small maintenance project consist of the following 10 jobs. Draw an arrow diagram
representing the project
1-2 2-3 2-4 3-5 3-6 4-6 4-7 5-8 6-8 7-8
4 6 10 8 2 12 4 15 14 18
226
Module V
OR SOFTWARES
Operations research uses various optimization algorithms to help make decisions related
to highly complex problem
Linear programming (LP) and mixed integrated programming (MIP) are often used to
solve these high complex decision making problem
LP maximizes (or minimize) a linear objective function subject to one or more constraint.
The technique finds broad use in OR and occasionally of use in statistically work
NCSS solves a particular linear programme using a revised dual simplex method
available in the extreme optimization mathematical subroutine package
This technique requires a special data format which will be discussed under the
specification tab
Using the analysis menu or the procedure navigator, find and select the linear
programming with bounds procedure.
On the menus, select file, then new template. This will fill the procedures with the default
temp plate
227
• Select type from the best of columns and then click ok. Type will appear in this
box
• Double click in the variable column text box. This will bring up the column
selection window
• Select X1 – X4 from the list of columns and then click ok : x1 – X4 will appear in
the box
• Double click in the labels constraints column text book . This will bring up the
column selection
• Select C label from the list of columns and then click ok. (Label will appear in this
box)
• Double click in the logic column textbook. This will bring up the column select
window)
• Select logic from the list of columns and then click ok. Logic will appear in this
box.
• Double click in the constraint bounds (RHS) columns text book. This will bring
up the various selection window.
• Select RHS from the list of column and then click Ok. RHS will appear in this
box.
228
Theory Questions Part A&part B questions in OR
229
a. .North West Corner Rule
b .Least cost entry method
c .Vogels Approximation method
d .MODIs test of optimality
27.State Unbalanced Transportation problem
28Explain degeneracy in transportation problem.How can be controlled
29.State the difference between Transportation problem and Assignment problem
30.What is meant by Simplex method. Describe the steps involved in solving Simplex
method
31.Write short notes on
A .Duality in Simplex Method
B .Artificial variables
C .Big M Method
32.Explain the special types of LPP problem
33.Write short notes on Degeneracy in Simplex Method
34.Write short notes on Sensitivity Analysis
35.State the objectives of Queuing Theory
36.What are the applications of Queing Theory
37.State the role of Queuing Theory in Management
38 Explain the types of Queuing system
39.Explain the customer behavior in queing system
40.State the characteristics of Queing Model
41.Explain with examples service cost for multiple channel system
42.Write short notes on Monte Carlo Simulation
43.Explain Traffic Intensity
44.Write short note on Kendels notation for queing
45.Explain Simulation modeling
46. Define Simulation Give two applications areas where Simulation Technique is
suitable
47.What are the uses of simulation
48.What is game theory? Explain the scope of game theory
49.What is meant by game
230
50.Write short note on Pay off matrix
51.What is meant by value of game
52. Explain the different types of game
53.what are the limitations of game theory
54.State Maxi Min Principle and Minimax principle used in game theory
55.Write short notes on Saddle point
56.Explain Mixed Strategy ?Explain how the solution of mixed strategy be determined
57.State Dominance Theory of value of game
58.What is decision theory
59.What is the significance of opportunity loss in decision making under risk
60.Distinguish between decision making under uncertainty and decision under risk
61.Explain the steps involved in decision theory
62.Write short notes on Opportunity loss
63.Discuss in detaild the different decision environments
64.Write short notes on criteria of Optimism and Pessimism
65.Explain Laplace criteria in Decision making
66.ExplainHurvizCriterian
67.What is meant by regret table
68.Write short notes on EMV &EOL
69.What is meant by expected value of perfect information
70. Explain Bayesian role of decision making approach
71.Explain the objectives of Net Work Analysis
72.ExplainNetworkAnalysis.What are the uses
73.Explain the phases of Project management
74.What is meant by float? Describe the different types of float
75.What are the uses of float
76.What is critical path method(CPM) Explain the steps involved in CPM
77.Write short notes on PERT
78.Differentiate between PERT and CPM
79. Explain the steps involved in construction of PERT
80 .State the applications of PERT
81 .Write short notes on crashing of Projects
231
82. Write short notes on Resource Smoothening
83.Write short notes on Resource leveling
84.Explain the terms Optimistic, Pessimistic ,most likely time estimates in project
Management
85.Describe the different soft wares used in Operations Research
86 .Critically comment on the assumption on which PERT/CPM analysis is done for
projects
87.Write short notes on Decision Tree Approach
88.Wrie short notes on Project variance in PERT
232