UNIT-1

Importance and Scope of Operations Research

Introduction

Operation is simply defined as the research of operations. Operations may be called a

set of acts required for the achievement of a desired outcome. Operation research is used to
maximize the utility of limited resources. The objective is to select the best alternative that is
the one leading to the best result. O.R is scientific approach to problem solving for executive
management

Scope of operations research in Management

The technique used in operations research has very wide application in various fields of
business/industrial/government/social sector. Few areas of applications are mentioned below:

Marketing and sales:

1. Product selection and competitive strategies

2. Utilization of salesmen, their time and territory control, frequency of visits in sales force
analysis.
3. Marketing advertising decision for cost and time effectiveness.
4. Forecasting and decision trends

Production Management

1. Product mix and product proportioning

2. Facility and production planning and scheduling sequencing
3. Physical distribution, warehousing and retail outlets planning-nature and localities
4. Project planning, scheduling , allocation of resources, monitoring and control systems
5. Queuing system design.
6. Quality control decision

Purchasing, procurement and inventory controls

1. Buying policies and prices

2. Negotiation and bidding policies
3. Time and quantity of procurement
Finance, Investments and Budgeting

1. Profit Planning
2. Cash flow analysis
3. Investment policy for maximum return
4. Dividend policies
5. Portfolio analysis

Personnel Management

1. Determination of optimum organizational level

2. Job evaluation and assignment analysis
3. Mixes of age and skills
4. Recruitment policies and job description

Research and Development

1. Determination of areas of thrust for research and development

2. Selection criteria for specific project
3. ‘what if’ analysis for alternative design and reliability
4. Trade –off analysis for time-cost relationship and control of development project.

Importance of Operations Research

Business Operations

OR could be very effective in handling issues of inventory planning and scheduling,

production planning, transportation, financial and revenue management and risk management.
Basically, OR could be used in any situation where improvements in the productivity of the
business are of paramount importance.

Control

With OR, organizations are greatly relieved from the burden of supervision of all the
routine and mundane tasks. The problem areas are identified analytically and quantitatively.
Tasks such as scheduling and replenishment of inventories benefit immensely from OR.

Decision Making

OR is used for analyzing problems of decision making in a superior fashion. The

organization can decide on factors such as sequencing of jobs, production scheduling and
replacements. Also the organization can take a call on whether or not to introduce new products
or open new factories on the basis of a good OR plan.
Coordination

Various departments in the organization can be coordinated well with suitable OR. This
facilitates smooth functioning for the entire organization.

Systems

With OR, any organization follows a systematic approach for the conduct of its business.
OR essentially emphasizes the use of computers in decision making; hence the chances of errors
are minimum.

Classification of OR Models
The first thing one has to do to use O.R. technique after formulating a practical problem
is to construct a suitable model to represent the a practical problem. A model is a
reasonably simplified representation of a real-world situration. It is a abstraction of
reality. The model can broadly be classified as

Iconic(Physical) Models

Analog Models

Mathematical Models

Also as

Static Models

Dynamic Models

Deterministic Models

Stochastic Models

Models can further be subdivided as

Descriptive Models

Prescriptive Models

Predictive Models

Analytic Models
Simulation Models

Iconic Model:
This is physical or pictorial representation of various aspects of a system.

Example: Toy, Miniature Model of a building, scaled up model of a cell in biology etc.

Analogue or Schematic Model:

This uses one set of properties to represent another set of properties which a system
under study has

Example: A network of water pipes to represent the flow of current in an electrical

network or graph, organizational charts etc.

Mathematical model or Symbolic model:

This uses a set of mathematical symbols (letters, numbers etc) to represent the decision
variables of a system consideration. These variables related by mathematical equations
or in equations which describes the properties of the system.

Example: A linear programming model, A system of equations representing an electrical

network or differential equations representing dynamic systems etc.

Static Model:

This is a model which does not take time into account. It assumes that the values of the
variables do not change with time during a certain period of time horizon.

Example: A linear programming problem, an assignment problem, transportation

problem etc.

Dynamic Model:

This is a model which considers time as one of the important variables.

Example: A dynamic programming problem. A replacement problem.

Deterministic model:

This is a model which does not take uncertainty into account.

Example: A linear programming problem, an assignment problem etc.

Stochastic model:
This is a model which considers uncertainty as an important aspect of the problem

Example: Any stochastic programming problem, stochastic inventory models etc,

Descriptive model:

This is one which just describes a situation or system.

Example: An opinion poll, any survey.

Predictive Model is one which predicts something based on some data. Predicting
election results before actually the counting is completed.

Prescriptive model is one which prescribes or suggests a course of action for a problem

Example: Any programming (linear, non linear, dynamic, geometric etc.) problem

Analytic model:

This is a model in which exact solution is obtained by mathematical methods in closed

form.

Simulation model: This is a representation of reality through the use of a model or

device which will react in the same manner as reality under a given set of conditions.
Once a simulation model is designed, if takes only a little time, in general , to run a
simulation on a computers.

Phases of O.R
1) Formulation of the problems: Identifying the objective, the decision variables
involved and the constraints that arise involving the decision variables.
2) Construction of a Mathematical Model: Expressing the measure of effectiveness
which may be total profit, total cost, utility etc., to be optimized by a Mathematical
function called objective function. Representing the constraints like budget
constraints, raw materials constraints, resource mathematical equations or
inequalities.
3) Solving the Model constructed: Determining the solution by analytic or iterative or
Monte-Carlo method depending upon the structure of the mathematical model.
4) Controlling and Updating
5) Testing the model and its solution i.e., validating the model: Checking as far as
possible either from the past available data or by expertise and experience
whether the model gives a solution which can be used in practice.
6) Implementation: Implement using the solution to achieve the desired goal.
 Linear Programming
Linear programming is a mathematical programming technique to optimize the
performance (profit or cost), under a set of resource constraints (Machine hours, money,
material, time men hours etc). As a special by an organization the usefulness of this technique is
maximize even though the competitions in a linear programming model are two elaborate a
sample list of applications of the linear programming problem is presented below.

1. Product mix problem

2. Diet planning problem
3. Manpower planning problem
4. Capital budgeting problem

Concept of Linear programming problem

The model of any linear programming problem will contain objective fuction, set of
constraints and non – ve restrictions. Each of the company may consist of one of more of the
following.

1. Decision variables
2. Objective function coefficients
3. Technological co efficient
4. Resource availabilities

Formulation of Linear programming problem (LPP)

A company manufactures two different types of product P1 and P2. Each type of product
requires processing of milling machine and drilling machine. Each type of machine as limited
hours available for week the net profit per unit of the product resource requirement of the
products of the products and availability of the sources are summarizing in table.

Machine type Processing time (in hours) Machine hours

P1 P2 Available per week
Milling machine 2 5 200

Drilling machine 4 2 240

Profit
(Per unit in Rs.) 250 400
Solution

Let ‘n’ be the no. of products to be manufacture and ‘m’ be the different machine types

Decision variables

The decision variable is used represent the levels of achievement of a particular course of
action. The solution of LPP will provide the optimal value for each and every decision variable
of the model a generalize definition of a decision variable is presented below

From the example ‘x1’ is the production volume of product ‘P1’. ‘x2’ is the production
volume of the ‘P2’ and ‘x1’ and ‘x2’ are called decision variables.

Objective function coefficients

It is a constant representing the profit or cost per unit of the caring out an activity.

Let C1, C2, …… Cn be the profit per unit of the products P1, P2, P3 … Pn respectively then
from the example C1 is the profit per unit of the product ‘P1’ which is equal to Rs. 250 and ‘C2’
be the profit per unit of the product ‘P2’. Which is equal to Rs.400 In this example ‘C1’ and ‘C2’
are called objective function coefficients.

Objective function

It is an expression representing the total profit or cost per caring our. A set of objectives
at some levels the objective function will be maximization or minimization. The benefit related
objective function will come under maximization. Whereas the cost related objective function
will come under minimization type a generalized format of the objective function is presented
below.

Min or Max Z = C1x1 +C2 x 2 +C3 x 3 +.........Cn x n

From the example

Max Z =250x1 + 400x2

Technological coefficients

The technological coefficients are the amount of resource ‘i’ required for activity ‘j’.
Where ‘i’ varies from 1, 2, 3 …n and j varies from 1, 2, 3, …… m. A generalized formal of the
technological coefficients matrix is
  a11 a12 ... a1n 
a 
 21 a22 ... a2 n  2 5
 . . ... .  From the example  
  4 2
. . ... . 
 am1 am 2 ... amn 

Where a11 = 2, a12 = 5, a21 = 4 and a22 = 2.

Resource Availability

The constraints ‘b’ is the amount of resource ‘i’ available during the planning period.

A generalized format of the resource availability matrix is

 b1 
b 
 2
.  b1   200 
  From the example   =  
. b2   240 
.
 
bn 

Set of constraints or subject to constraints

A constraint is kind of restriction and the total amount of a particular resource required to
carry out at various levels in a model there will be many such constraints so the constraints will
reveal the levels of achievement different decision variables. A generalized form of the
constraints can be given as

a11 x1 + a12 x2 + ...........a1n xn  or  b1

a21 x1 + a22 x2 + ...........a2 n xn  or  b2
a31 x1 + a32 x2 + ...........a3n xn  or  b3
.
.
.
am1 x1 + am 2 x2 + ...........amn xn  or  bn
From the example

a11 x1 + a12 x2  b1
a21 x1 + a22 x2  b2
substitute the value of a11 , a12 , a21 , a22, b1 and b2
2x1 + 5 x2  200
4 x1 + 2 x2  240

Non – negative Restrictions

Each and every decision variable in the linear programming model is a non negative
variable. The condition is represents as complete model this example

Max z =250x1 + 400 x2

Set of constraints
2x1 + 5 x2  200
4 x1 + 2 x2  240
x1 , x2  0

Properties of Linear Programming model

Optimum solution

If there is no other superior solution to the solution obtained for a given linear
programming model then the solution is obtained a treated as the optimum solution.

Feasible solution

If all the constraints of the linear programming model are satisfied by the solution of the
model then the solution is called feasible solution.

Infeasible solution

When the constraints are not satisfied simultaneously, the LPP has no feasible solution.
That solution is called infeasible solution. This means that there is no solution for the given
model which can be implements.

Unbounded solution

When the values of the decision variable may be increased indefinitely or infinitely
without violating any of constraints, the solution space (feasible region) is unbounded. The value
of objective function in such cases, may increase (for maximization) or decrease (for
minimization) indefinitely. Thus both solution space and objective function value are
unbounded.
Bounded solution

When the values of the decision variable may by definite or finite with violating any of
constraints, the solution space (feasible region) is bounded.

Linear Programming Problem Methods

Algorithm for Graphical Method:

The major steps in the solution of a linear programming problem by graphical method are
summarized as follows:

Step 1: identify the problem – the decision variables, the objective and restrictions.

Step 2: Set up the mathematical formulation of the problem.

Step 3: Plot a graph representing all the constraints of the problem and identify the feasible
region (solution space). The feasible region is the intersection of all the regions represented by
the constraints of the problem and is restricted to the first quadrant only.

Step 4: The feasible region obtained in “step – 3” may be bounded or unbounded. Compute the
co-ordinates of all the corner points of the feasible region.

Step 5: Find out the value of the objective function at each corner (solution) point determined in
“step -4”.

Step 6: Select the corner point that optimizes (maximizes or minimizes) the value of the
objective function. It gives the “optimum feasible solution”.

Slack Variables:

If the constraints has ≤ type in order to make it in equation we have to add

something +ve quantity to left hand side. +ve quantities are called slack variables
Ex:

2 x1 + 4 x2  3
3 x1 + 5 x2  4
convert them into equations
2 x1 + 4 x2 + s1 = 3
3 x1 + 5 x2 + s2 = 4
where s1 and s 2 are slack variables

Surplus variables

If the constraints have ≥ type in order to make it in equation we have to add

something -ve quantity to left hand side. -ve quantities are called surplus variables.

Ex:

2 x1 + 4 x2  3
3x1 + 5 x2  4
convert them into equations
2 x1 + 4 x2 − s1 = 3
3x1 + 5 x2 − s2 = 4
where s1 and s 2 are surplus variables

Algorithm for Simplex Method:

Step 1: Check whether the given objective function is maximization type

Case a: If it is maximization type then go to next step.

Case b: If is minimization problem then we converting them into maximization type by

using relation.

Min Z = -Max Z*

Step 2: check whether the right hand side constants are non – negative (positive)

Case a: if is non – negative then go to next step.

Case b: if at least one of the constant is negative (-ve) then we converting into +ve value
by multiplying the corresponding constraint by “-1”.
Step 3: Introduce slack and surplus variables to the constraints and the cost of objective function
becomes zero.

Step 4: Express the given objective functions and constraints in matrix form.

Step 5: Check whether there exist initial basic feasible solution (IBFS).

Step 6: prepare initial simplex table by using the following columns.

Cj C1 C2 C3……………………Cn
CB SB XB X1 X2 X3…………………… Xn S1 S2 S3 …………… Sn

Zj –Cj

Step 7: Find z-value i.e., z = CB XB

Step 8: Find net evaluations i.e., Zj – Cj where j = 1, 2, 3, ……..n

Step 9: Check whether all the net evaluations are non –ve.

Case a: if it is so then there exist optimum solution.

Case b: If at least one net evaluation is –ve then go to next step

Step 10: Find entering variable which is most –ve net evaluation i.e., Min −( Z j − C j )

Step 11: Find leaving variable which is

 XB 
Min   , Entering column elements > 0.
 Entering column elements 

Step 12: Find pivotal or key element which is intersection of both entering and leaving variables.

Step 13: Do the simplex operations, new values of

Old values of the key row

Key row =
Key element

New values of the Non key row = Old values of the Non key row - (Non key element)  (New values of the key row)
Step 14: Repeat the procedure. Step 5 to step 12. Until optimum solution are unbounded solution
has been obtained.

Big – M or Mixed or Penalty Method or Alternate Method:

Step 1: Check whether the given objective function is maximization type.

Case a: If it is maximization then go to next step

Case b: If it is minimization type then we convert them into maximization type by using
relation.

Min Z = -Max Z*

Step 2: Check whether the right hand side constants are non –ve

Case a: If it is non –ve or positive then go to next step.

Case b: If it is –ve or negative then we convert them into positive (+ve) by multiplying
the corresponding constraint by “-1” Automatically the signs will also change.

Step 3: Express the given objective function and constraints in matrix form.

Step 4: Check whether there exist IBFS

Case a: If it is exit so then go to next step

Case b: If it is not so, then introduce artificial variable to the constraints ≥ type and the
cost co-efficient objective function becomes “-M”.

Step 5: Prepare Initial Simplex Table

C1 C2 C3………Cn
Cj
CB SB XB X1 X2 X3…… Xn S1 S2 S3 …..Sn A1,A2,…………An

Zj –Cj

Step 7: Find Z-value Z =  CB X B .

Step 8: Find Net evaluations i.e., Z j − C j (j = 1, 2, 3 ….n)

Step 9: Check whether the net evaluations are non –ve.

Case a: If it is so and verifies the following conditions.

i) If all net evaluations are non –ve. If at least one artificial variable.
Still in the basis not at zero level then there exist infeasible solution.
ii) If all net evaluations are non –ve and if at least one artificial variable still in the
basis at zero level then there exist optimum solution.
iii) If all net evaluations are non –ve there is no artificial variable in the basis then
there exist optimum solution.

Case b: If it is so then go to next step.

Step 10: Find entering variable which is most –ve net evaluation i.e., Min −( Z j − C j )

Step 11: Find leaving variable which is

 XB 
Min   , Entering column elements > 0.
 Entering column elements 

Step 12: Find pivotal or key element which is intersection of both entering and leaving variables.

Step 13: Do the simplex operations, new values of

Old values of the key row

Key row =
Key element

New values of the Non key row = Old values of the Non key row - (Non key element)  (New values of the key row)

Step 14: Repeat the procedure. Step 7 to step 13. Until optimum solution are unbounded solution
has been obtained.

Duality in Linear programming problem

Every LPP there always exist another LPP which is basing on the same data and having
same solution. The original problem is called “Primal problem” the associated one is called
“Dual Problem”.
Rules for Converting Primal problem Into Dual:

If the system of constraints in a given LPP consist of mixed equations (or ) in equations
non –ve variables or unrestricted variables then the dual of the given problem can be obtained
by reducing to standard primal problem.

Step 1: first convert the objective function to maximization form, if not.

Step 2: If the constraints as ≥ type then we converts them into ≤ type by multiplying -1.

Step 3: If the constraints as an equal sign then it is replaced by two constraints involving the in
equalities is in opposite direction simultaneously.

Step 4: Every unrestricted variables is replaced by difference of “2” Non –ve variable.

Step 5: We get the standard primal form of the given LPP in which

Case a: All the constraints have ≤ type the objective function is maximization.

Case b: All the constraints have ≥ type when the objective function is minimization.

Step 6: Identify the variables to be used in the dual problem. The no. of these variables equals to
the no. of constraints in the primal problem.

Step7: Write the objective function. Objective function coefficient is the Right hand side
constraints of primal constraints. If the primal problem is of maximization type the dual will be
minimization problem and vice versa.

Step 8: Making use of “Step 6” write the constraints for the dual problem

Case a: If the primal is a maximization problem the dual constraints must be all are ≥
type, If the primal is minimization problem the dual constraints must be all are ≤ type.

Case b: The column co-efficient of the primal constraints become the row coefficients of
dual constraints.

Case c: The co-efficient the primal objective function become the right hand side
constants of the dual constraints.
 Applications of Linear Programming
Linear programming is the most widely used techniques of decision –making in business
and industry and in various other fields. In this section, we will discuss a few of the broad
application areas of linear programming.

Agricultural Application:

This application fall into categories of farm economics and farm management. The farm
deals with agricultural economy of a nation or region.

Linear programming study of farm economics deals with interregional competition and
optimum allocation of crop production.

Linear programming can be applied in agricultural planning e.g. allocation of limited

resources such as acreage, labor, water supply and working capital, etc. in a way so as to
maximize net revenue.

Production Management:

Product Mix: A company can produce several different products each of which requires the use
of limited production resources. In such cases, it is essential to determine the quantity of each
product to be produced knowing its marginal contribution and amount of available resources
used by it. The objective is to maximize the total contribution, subject to all constraints.

Production planning: This deals with determination of minimum cost6 production plan over
planning period of an item with a fluctuating demand, considering the initial number of units in
inventory, production, man power and all relevant cost factors.

Financial Management

Portfolio Selection: This deals with the selection of specific investment activity among several
other activities. The objective is to find the allocation which maximizes the total expected return
or minimize risk. Under certain limitations.

Profit planning: This deals with the maximization of the profit margin form investment in plant
facilities and equipment, cash in hand and inventory.

Marketing management:

Medial selection: Linear programming techniques help in determining the advertising media
mix. So as to maximize the effective exposure, subject to limitation of budget, specified

exposure rates. To different market segments, specified minimum and maximum no. of
advertisements in various media.
 UNIT-2
Transportation Problem

Introduction:

The transportation problem is one of the subclasses of LPPs in which goods are
transported from a set of sources to a set of destinations subject to the supply and demand of
the source and destination, respectively, such that the total cost of transportation is minimized.

Examples of Transportation Problem

Source Destination Commodity Objective

Plants Markets Finished goods Minimizing total cost

of shipping

Plants Finished goods Finished goods Minimizing total cost

warehouses of shipping

Suppliers Raw material Raw material Minimizing total cost

warehouses of shipping

The solution of transportation problem can be obtained in two stages, namely initial solution or
initial basic feasible solution and optimum solution.

Initial solution
Initial solution cam be obtained by using any one of the three methods viz.,

I. North West corner rule (NWCR).

II. Row minimum rule
III. Column minimum rule.
IV. Least cost method or Matrix minima method.

1
 V. Vogel’s approximation method (VAM).

VAM is preferred over the other two methods, since the initial basic feasible solution obtained
by this method is either optimal of very close to the optimal solution.

Optimum solution
The improved solution of the initial basic feasible solution is called optimal solution which is the
second stage of solution that can be obtained by MODI (modified distribution method) or U-V
method.

Algorithm for Northwest-Corner rule

Step 1 : Find the minimum of the supply and demand values with respect to the current
northwest corner cell of the cost matrix.

Step 2 : Allocate this minimum value to the current northwest-corner cell and subtract this
minimum from the supply and demand values with respect to the current northwest-corner cell.

Step3 : Check whether exactly one of the row/column corresponding to the northwest-corner
cell has zero supply/demand, respectively. If so, go to step 4.

Step4 : Delete that row/column with respect to the current northwest-corner cell which has the
zero supply/demand and go to next step.

Step5 : Repeat above steps moving down towards the lower right corner of the transportation
table until all the rim requirements are satisfied.

Algorithm for Least Cost or Matrix Minima Method

Step 1 : Find the minimum of the values in the cost matrix (i.e. find the matrix minimum).

Step 2 : Find the minimum of the supply and demand values(X) with respect to the cell
corresponding to the matrix minimum.

Step 3 : Allocate X units to the cell with the matrix minimum. Also, subtract X units from the
supply and the demand values corresponding to the cell with the matrix minimum.

Step 4 : Check whether exactly one of the row/column corresponding to the cell with the matrix
minimum has zero supply/demand, respectively. If yes, go to step 5.

Step 5 : Delete that row/column with respect to the cell with the matrix minimum which has the
zero supply/zero demand.

Step 6 : Repeat above steps for the resulting reduced transportation table until all the rim
requirements are satisfied.

2
Algorithm for Vogel’s Approximation Method (VAM)
Step 1 : Find row penalties, i.e. the difference between the first minimum and the second
minimum in each row . If the two minimum values are equal, then the row penalty is zero.

Step 2 : Find column penalties, i.e. the difference between the first minimum and the second
minimum in each column. If the two minimum values are equal, then the column penalty is zero.

Step 3 : Find the maximum amongst the row penalties and column penalties and identify
whether it occurs in a row or in a column. If the maximum penalty is in a row, go to step 4
otherwise go to step 7.

Step 4 : Identify the cell for allocation which has the least cost in that row.

Step 5 : Find the minimum of the supply and demand values with respect to the selected cell.

Step 6 : Allocate this minimum value to that cell and subtract this minimum from the supply and
demand values with respect to the selected cell and go to step 10.

Step 7 : Identify the cell for allocation which has the least cost in that column.

Step 8 : Find the minimum of the supply and demand values with respect to the selected cell.

Step 9 : Allocate this minimum value to the selected cell and subtract this minimum from the
supply and demand values with respect to the selected cell.

Step 10 : Check whether exactly one of the rows and the column corresponding to the selected
cell has zero supply/zero demand, respectively. If yes go to step 11.

Step 11 : Delete the row/column which has the zero supply/zero demand and revise the
corresponding row/column penalties.

Step 12 : Repeat the procedure until all the rim requirements are satisfied.

Algorithm for MODI (modified distribution method) or U-V method.

Step 1 : Find the initial basic feasible solution of a Transportation Problem by using any one of
the above methods.

Step 2 : Check whether the number of basic cells (occupied cells) in the set of initial basic
feasible solution is equal to m+n-1. If yes go to next step.

Step 3 : Compute the values for U1, U2, U3, ………. Um and V1, V2, V3, V4, …………. Vn by applying the
following formula to all the basic cells only.

3
 Ui + Vj = cij

Step 5 : Compute penalties Pij or the non-basic cells by using the formula.

Pij = Ui + Vj - cij

Step 6 : Check whether all Pij values are less than or equal to zero . If yes, go to step 11,
otherwise, go to step 7.

Step 7 : Identify the non-basic cell which has the maximum positive penalty, and term that cell
as the new basic cell.

Step 8 : Starting from the new cell, draw a closed loop consisting of only horizontal and vertical
lines passing through some basic cells.(Note : Change of direction of the loop should be with 90
degrees only at some basic cell.)

Step 9 : Starting from the new basic cell, alternatively assign positive(+) and negative (-) signs at
the corners of the closed loop.

Step 10: Find the minimum of the allocations made amongst the negatively signed cells.

Step 11 : The optimality is reached. Treat the present allocations to the set of basic cells as the
optimum allocations.

Assignment Problem
Introduction

The objective of assignment problem is to assign a number of origins (jobs) to the equal number
of destinations (persons) at a minimum time or minimum cost or maximize the profit.

Algorithm for Assignment problem

Step 1: Determine the cost table from given problem

I. If the number of source is equal to the number of destinations, go to step 3

II. If the number of sources is not equal to the number of destinations, go to step 2

Step 2: Add a dummy source or dummy destination, so that the cost table becomes a square
matrix. The cost entries of dummy source/destinations are always zero.

Step 3: locate smallest element in each row of the given cost matrix and then subtract the same
from each element of that row.

4
Step 4: in the reduced matrix obtained in step3, locate the smallest element of each column and
then subtract the same from each element of that column. Each column and row now have at
least one zero.

Step 5: in the modified matrix obtained in step4, search for an optimal assignment as follows:

a. Examine the rows successively until a row with a single zero is found. Enrectangle this
zero (□) and cross off (x) all other zeros in its column. Continue in its manner until all the
rows have been taken care of.
b. Repeat the procedure of each column of the reduced matrix.
c. If a row and/or column has two or more zeros and one cannot be chosen byt inspection,
then assign arbitrary any one of these zeros and cross off all other zero of that
row/column.
d. Repeat (I) through (III) above successively until the chain of assigning (□) or cross (x)
ends. Solution is reached.

Step 6: if the number of assignments (□) is equal to n (the order of matrix) , an optimum
solution is reached

If the number of assignment is less than n (the order of matrix), go to next step.

Step 7: Draw the minimum number of horizontal and/or vertical lines to cover all the zeros of
the reduced matrix. This can be conveniently done be using a simple procedure:

a. Mark (√) rows that do not have any assigned zero.

b. Mark (√) columns that have zeros in the marked rows.
c. Mark (√) rows that have assigned zeros in the marked columns.
d. Repeat (b) and (c) above until the chain of making is completed.
e. Draw lines through all the unmarked rows and marked columns. This gives us the
desired minimum number of lines.

Step 8: Develop the new revised cost matrix as follows:

a. Find the smallest element of the reduced matrix not covered by any of the lines.
b. Subtract this element form all the unmarked elements and add the same to all the
Elements lying at the intersection of any two lines.

Step 9: Go to step 6 and repeat the procedure until an optimum solution is attained.

5
 Sequencing problems
Introduction:
The selection of an appropriate order for a series of jobs to be done on a finite
number of service facilities, in some pre-assigned order, is called sequencing.

A practical situation may corresponds to an industry producing a number of

products, each of which is to be processed through different machines, of course,
finite in number

The general sequencing problem may be defined as: Let there be n jobs to be
performed one at a time on on each of m machines. The sequence (order) of the
machines in which each job should be performed is given. The actual or
expected time required by the jobs on each of the machines is also given. The
general sequencing problem, therefore, is to find the sequence out of (n!)m
possible sequences which minimize the total elapsed time between the start of
the job in the first machine and the completion of the job on the last machine.

Assumptions:
• Each job once started on a machine, is to be performed up to completion
on that machine.

• The processing time on each machine is known. Such a time is

independent of the order of the jobs in which they are to be processed.

• The time taken by each job in changing over from one machine to another
is negligible.

• A job starts on the machine as soon as the job and the machine both are
idle and job is next to the machine and the machine is also next to the job.

• No machine may process more than one job simultaneously.

• The order of completion of job has no significance, i.e. no job is to be

given priority. The order of completion of jobs is independent of sequence
of jobs.

Basic Terms:
Number of Machines: It refers to the no. of service facilities through which a job
must pass before it is assumed to be completed.

6
Processing Order: It refers to the order (sequence) in which given machines are
required for completing the job.

Processing Time: It indicates the time required by a job on each machine.

Total Elapsed Time: It is the time interval between starting the 1 st job and
completing the last job including the idle time in a particular order by the given
set of machines.

Idle time on a machine: It is the time for which a machine does not have a job
to process, i.e. idle time from the end of job (i-1) to the start of jobs i.

Processing Time: It indicates the time required by a job on each machine.

No passing rule: It refers to the rule of maintaining the order in which jobs are to
be processed on given machines.

Process ‘n’ jobs through two Machines:

Optimum Sequence of Algorithm:

1. List the jobs along with their processing times in a table.

Job Number : J1 J2 J 3 .......... J n

M1 : t 11 t 12 t 13 .......... t 1n
ProcessingTime on machine 
M 2 : t 21 t 22 t 23 .......... t 2n

2. Examine the rows for processing times on machines M1 and M2, and find
the smallest processing time in each row, i.e. find out min.(t1j, t2j) for all j.
3. If the smallest processing time is for the first machine M1, then place the
corresponding job in the 1st available position in the sequence, otherwise
place 2nd machine M2.
4. If there is a tie in selecting the minimum of all the processing times, then
there may be 3 situations
a) Minimum among all processing times is same for the machines, i.e.,
min.(t1j, t2j)= t1k = t2r, then process the kth job first and the rth job last.
b) If the tie for minimum occurs among processing times t 1j on machine M1
only, then select arbitrarily the job to process first.
c) If the tie for minimum occurs among processing times t2j on machine M2,
then select arbitrarily the job to process last.

7
 d) Cross off the jobs already assigned and repeat steps 1 through 4, placing
the remaining jobs next to first or next to last, until all the jobs have been
assigned
➢ Calculate idle time for machines M1 and M2:
Idle time for M1 = Total elapsed time –(time when the last job in a
sequence finishes on M1.
Idle time for M2= Time at which the 1st job in a sequence finishes on
M1 + (time when the jth job in a sequence starts on M 2) – {(time
when the (j-1)th job in a sequences finishes on M2)}

➢ The total elapsed time to process all jobs through 2 machines as

under:

Total elapsed time = Time when the nth job in a sequence finishes on machine M 2

n n
=  t 2j +  I 2j
j= 2 j= 2

Where, t 2j = Time required for processing jth job on machine M 2

I 2j = Time for which machine M 2 remains idle after processing (j - 1)th
job and before starting work on jth job.

Process ‘n’ jobs through ‘K’ Machines:

Let there be ‘n’ jobs, each of with is to be processed through ‘m’ machines, says
M1, M2, M3,….. Mk, in order M1, M2, M3,….. Mk,. the list of jobs with their
processing times is:

Job number 1 2 3 ........... n

M1 : t11 t12 t13 ........... t1n
M2 : t21 t22 t23 ........... t2 n
Machines . . . . . t3 n
. . . . . t3 n
. . . . . t3 n
Mk : tk 1 tk 2 tk 3 ........... tkn

8
Optimum Sequence of Algorithm:
The iterative procedure for determining the optimal sequence for ‘n’ jobs on ‘k’
machines can be summarized as follows:

Step 1: Find Min t1j, Min tkj and Maximum of each of t2j,t3j,….tk-1j for all j = 1,2,…n

Step 2: Check the following

Min ( t1 j )  Max ( tij ) for i = 2,3,4....k-1

or
Min ( tkj )  Max ( tij ) for i = 2,3,4....k-1

Step 3: If the inequalities of step 2 are not satisified, method fails. Otherwise go
to next step.

Step 4: Convert the k machine problem into two-machine problem by introducing

tow fictitious machine G and H, such that

G = t1 j + t2 j + t3 j ....tk −1 j
H = t2 j + t3 j + t4 j ....tkj
Step 5: Determine the optimal sequence for the ‘n’ jobs and ‘2’ machines equitant
sequencing problems with the prescribed order GH in the same way as
discussed earlier the resulting sequence shall be optimum for the given problem.

Processing Two Jobs through m machines

Example 1. Use graphical method to minimize the time needed to process the
following jobs on the machines shown below also calculate the total time needed
to complete both the jobs.

Job1 A B C D E
Job 2 C A D E B

Solution:

Step -1: First draw a set of axes, where the horizontal axis represents processing
time on job 1 and the vertical axis represents processing time on job 2.

Step 2: layout the machine time for two jobs on the corresponding axes in the
technological order machine A takes 2 hours for job 1 and 5 hours for job 2.

9
Construct the rectangular PQRS for the machine A. Similarly, other rectangular
for machine B,C,D and E are constructed as shown in fig.

Step 3: make a program by starting from the origin O and moving through
various stages of Completion (points) until the point marked finish is obtained.
Physical interpretation of the path thus chosen involves the series of segments
which are horizontal or vertical or diagonal making an angle of 450 with the
horizontal. Moving to the right means that job1 is proceeding while job 2 is idle
and moving upward means that job 2 is proceeding while job 1 is idle and moving
diagonally means that both the jobs are proceeding simultaneously. Further both
the jobs cannot be processed simultaneously on the same machine. Graphically
diagonal movement through the
blocked out (shaded) area is not allowed and similarly for other machines too.

Step 4: (To find an optimal path) An optimal path is one that minimizes idle for
job1 (horizontal) and job 2 (vertical movement). Choose such a path on which
diagonal moment is as much as possible.

Step 5: (to find the elapsed time). The elapsed time is obtained by adding the
idle time either of the job to the processing time for that job. In this problem the
idle time for the chosen path is seen to be 3hrs for the 1 and 0 for the job 2. Thus
the total elapsed time, 17 + 3 = 20hrs

10
 UNIT-3
Replacement problems

Introduction: Replacement problems `are concerned with the situation that arise
when some items such as men, machines, electric-light bulbs, etc. Need replacement
due to their decreased efficiency, failure or break down. The deteriorating efficiency or
complete breakdown may be either grzadual or sudden,

Broadly speaking, the requirement of a replacement may be in any of the following

situations:

(i) An item fails and does not work at all or the item is expected to fail shortly.
(ii) An item deteriorates and need expensive maintenance.
(iii) A better design of the equipment is available.
(iv) It is economical to replace equipment in anticipation of costly failure.

Failure Mechanism of Items

The term ‘failure’ has a wider meaning in business than what it has in our daily life.
There are two kinds of failure.

1) Gradual Failure: The machanis under this category is progressive. That is, as
life of an item increases, its efficiency deteriorates causing.
a) Increased expenditure for operating costs,
b) Decreased productivity of the equipments,
c) Decrease in the value of equipment, i.e., the resale of saving value decrease
For example: Mechanical items like bearings, rings etc. Another example is
‘Automobile tires,
2) Sudden Failure: This type of failure is applicable to those items that do not
deteriorate markedly with service but which ultimately fail after some period of
using. The period between installation and failure is not constant for any
particular type of equipment but will follow some frequency distribution which
may be progressive, retrogressive or random in nature.

Replacement Policy when money value is not counted with time:

The aim here is to determine the optimum replacement age of an equipment/item

whose running/maintenance cost increases with time and the value of money
remains static during that period. Let

C: Capital Cost of equipment

S: Scrap value of equipment

n: Number of years that equipment would be in use.

f(t) : Maintenance cost functions,

A(n): Average total annual cost.

Case I: When time (t) is a continuous variable: If the equipment is used for ‘n’ years,
then the total cost incurred during this period is given by
TC = Capital cost − Scrap value+Maintance Cost
n
=C − S +  f ( t )dt
0

Average annual total cost, therefore is

TC
A( n) =
n
1 
n
=  C − S +  f ( t )dt 
n 0 
C −S 1 
n
= +   f ( t )dt 
n n0 
d
For minimum cost, we must have  A ( n )  = 0
dn 
d C − S 1  
n

 +   f ( t )dt   = 0
dn  n n0  

(V ) 
d d d
 Q Using (UV ) = V (U ) + U
d C−S  d 1 
n
dn dn dn
  +    f ( t )dt  =0  n

dn  n  dn  n  0   1 
 U = n and V =  f ( t )dt 
 0 
C−S  1
n
1   d n  
=− +− 2  f ( t )dt + f ( n )  =0  Q   f ( t )dt  = f ( n ) 
n2  n 0
n   dn  0  
C−S 1
n
1
=
n2
+ 2
n  f ( t )dt − n f ( n ) =0
0
 C−S 1
n
1
f (n) = + 2  f ( t )dt
n n2 n 0

C−S 1
n
f (n) = +  f ( t )dt
n n0
f ( n) = A( n)
d2
 A ( n )   0 at f ( n ) = A ( n )
dn 2 
i.e., when the maintenance cost becomes equal to the average annual cost, the
decision should be to replace the equipment.

Case 2: when time ( t ) is a discrete variable:

Here, the period of time is considered as fixed and n,
time ( t ) take the values 1,2,3... Then
C−S 1 n
A( n) = +  f (t )
n n i =0
Now, A ( n ) will be a minimum for that value of n, for which
A ( n + 1)  A ( n ) and A ( n − 1)  A ( n ) or
A ( n + 1) − A ( n )  0 and A ( n − 1) − A ( n )  0
For this, we write
C−S 1 n +1
A ( n + 1) =
n +1
+ f
n + 1 i =0
(t )
C−S 1  n 
=
n +1
+
n +1
f
 i =0
(t ) + f ( n + 1) 

C−S 1  n  1
=
n +1
+
n +1
f
 i =0
(t )  +
 n +1
f ( n + 1)
1  n
 1
=
n +1

C−S +  f ( t )  + n + 1 f ( n + 1)
i =0
  n

=
1
n +1
( nA ( n ) ) +
1
n +1
f ( n + 1)  Q nA ( )
n = C − S +  f (t ) 
 i =0 
We subtract A ( n ) on both sides

A ( n + 1) − A ( n ) =
1
n +1
( nA ( n ) ) +
1
n +1
f ( n + 1) − A ( n )

=
1
n +1
( f ( n + 1) − A ( n ) )
A ( n + 1) − A ( n )  0
1
n +1
( f ( n + 1) − A ( n ) )  0
f ( n + 1) − A ( n )  0
f ( n + 1)  A ( n )
Similarly it can shown that
A ( n ) − A ( n − 1)  0  f ( n )  A ( n − 1)

This suggest the optimal replacement policy:

Replace the equipment at the end of n years, if the
( n + 1)
th
maintenance cost in the year is more than
the average cost in the n th year and the n th year's
maintenance cost is less than the previous year's
average total cost.

Replacement Policy when Value of Money is counted with time:

In this case, the investor is paying interest on the money that has been invested. We
assume that
(i) The maintenance costs are incurred at the beginning of time periods; and
(ii) The maintenance costs increase with time.

Let the money carry an interest rate of r% per year, i.e., one rupee in n years time is
equivalent to Rs. (1+r)-n today.

(1+r)-n is called the present worth factor of rupee one spent n years after now.

(1+r)n is the payment compound amount factor of rupee one spent in n years time.
Let
C : Initial cost of equipment
R : Running cost of the equipment in year .
V : Rate of interest or Present worth factor
V = (1 + r )
−1

Vn : Future discounted costs associated with

the policy of replacing the equipment at
the end of each n years.
n −1
C + V k Rk
Vn = k =0

1−V n
Now Vn is minimum
Vn +1  Vn and Vn −1  Vn
n
C + V k Rk
Vn +1 − Vn = k =0
− Vn
1 − V n +1
C + V k Rk − Vn (1 − V n +1 )
n

= k =0

1 − V n +1
 n −1


 C + 
n −1
V k
R

k  + V R n − Vn (1 − V
n n +1
)  C +  V k
R k 
 k =0   Q Vn = k =0

=  1−V n

1 − V n +1  n −1 
( − )n = + 
n k
1 V V C V Rk 
 k =0 
=
(1 − V )V
n
n + V n
R n
− Vn (1 − V n
V )
1 − V n +1
Vn − V nVn + V n Rn − Vn + Vn  V n  V
=
1 − V n +1
V n ( Rn − Vn + Vn  V )
=
1 − V n +1
V n ( Rn − (1 − V )Vn )
=
1 − V n +1
Vn +1 − Vn  0
V n ( Rn − (1 − V )Vn )
0
1 − V n +1
V n ( Rn − (1 − V )Vn )  0
( R − (1 − V )V )  0
n n

Rn  (1 − V )Vn − − − − − − − − − − − − − − − − (1)
and similarly
V n −1 ( Rn −1 − (1 − v )Vn )
Vn − Vn −1 =
1 − V n −1
Vn − Vn −1  0
V n −1 ( Rn −1 − (1 − V )Vn )
n −1
0
1−V
Rn −1  (1 − V )Vn − − − − − − − − − − − − (2)
From (1) and ( 2 )
Rn −1  (1 − V )Vn  Rn
Replacement of items that fail completely:
Some situations, failure of a certain item occurs all of a sudden, instead of gradual
deterioration (e.g., failure of light bulbs, tubes, etc.).
The failure of the item may result in complete breakdown of the system. The breakdown
implies loss of production, idle inventory, idle labour, etc. Therefore, an organization
must prepare itself against these failures.

Thus, to avoid the possibility of a complete breakdown, it is desirable to formulate a

suitable replacement policy. The following two courses can be followed in such
situations.

• Individual replacement policy. Under this policy, an item may be replaced

immediately after its failure.
• Group replacement policy. Under this policy, the items are replaced in group
after a certain period, say t, irrespective of the fact that items have failed or not. If
any item fails before its preventive replacement is due, then individual
replacement policy is used.
 UNIT-4
GAME THEORY

Introduction:

A competitive situation will be called a “Game”. Many practical problems require

decision-making in a competitive situation where there are two or more opposing parties
where the action of one depends up on the one taken by the opponent.

For example: candidates for an election advertising and marketing campaigns by competing
business firms, countries involved in military bottles etc.

If Game has the following properties

➢ Each player has all relevant information available

➢ Each Game has finite number of players
➢ Each player can use finite number of strategies.

Some Basic Terms:

Player: the participants or competitors in the game are known as players. A player may be
individual or group of individuals.

Strategy: A strategy for a player is defined as a set of rules of alternative courses of action
available to him in advance, by which player decides the course of action that he should adopt
strategy may be of two types.

➢ Pure strategy
➢ Mixed strategy

Pure strategy: If the player select the game strategy each time , then it is referred to as pure-
strategy. The objective of the players is to maximize gains or to minimize losses.

Mixed strategy: When the player use a combination of strategies and each player always kept
guessing as to which course of action is to be selected by the other player at a particular
occasion then this is known as mixed strategy. There is a probabilistic situation and objective of
the player is to maximize expected gains or to minimize losses.

Optimum strategy: A course of action or play which puts the player in the most preferred
position, irrespective of the strategy of his competitors is called an optimum strategy.

1
Pay-off matrix:

When the players select their particular strategies. The pay-off (gains of losses) can be
represented in the form of a matrix called the payoff matrix.

Player B
 a11 a12 a13 
Player A a 21 a 22 a 23 
 a 31 a 32 a 33 

Let player ‘A’ have ‘m’ strategies A and player ‘B’ have ‘n’ strategies. It is assume that
each player has his choices from amongst the pure strategies. Also it is assumed that player ‘A’
is always the gainer and player ‘B’ is always looser.

Value of the Game: It is the expected pay off of play when all the players of the game follow
their optimum strategies. The game is called value of the game

Optimum strategy:

A course of action or play which puts the player in the most preferred position
irrespective of the strategy of his competitors is called on optimum strategy.

Value of the game:

It is the expected pay off of play when all the players of the game follow their optimum
strategies. The game is called fair is the value of the game is zero and unfair if it is non-zero.

Saddle Point:

The definition of pure strategy states that both the players use the same strategy every
time and that is the best strategy from the points of both players. In such case, the strategy
indicates the saddle point, which is the point maximum gain for A and minimum loss for B.

Rules for determine a saddle point:

We may now summarize the procedure of locating the saddle point of a pay off matrix
as follows:

Step 1: select the minimum element of each row of the pay off matrix and mark them (*)

2
Step 2: select the maximum element of each column of the pay off matrix and mark them (+)

Step 3: if there appears an element in the pay off matrix in the pay off matrix marked (*) and (+)
both, the position of that element is a saddle point of the pay off matrix

Methods form solving game without saddle points

When there are no saddle points in the game, pure strategies as optimal strategies are
not available. Such games can be resolved by using mixed strategy and following methods can
be adopted

1. Algebraic method.
2. Graphic method
3. Dominance method.
4. Linear programming method

Algebraic method:

Algebraic formula for matrix method:

The general formulae can be derived algebraically from the matrix in the following
manner. The matrix so obtained from rule of dominance can be written as:

Player B
a a12  p1
Player A  11 
 a 21 a 22  p2
q1 q2
Strategy for Player A:

A A2 
SA =  1
 p1 p2 
a22 − a21
p1 =
(a22 + a11 ) − (a12 + a21 )
p2 = 1 − p1

3
Strategy for Player B:

A A2 
SB =  1
 q1 q2 
a22 − a12
q1 =
(a22 + a11 ) − (a12 + a21 )
q2 = 1 − q1

Value of the game

(a22  a11 ) − (a12  a21 )

V=
(a22 + a11 ) − (a12 + a21 )

Where P = probability or chance of player A using strategy and q of player B using strategy.

Graphical Method:

Some games will be specialized nature like 2  n or m 2 m 2 games. The pay off matrix
for the 2 n games will contain 2 rows and n columns, where as the pay off matrix of the m 2
game will contain ‘m’ rows and 2 columns. If there is no saddle point for these games, one can
solve them using graphical method.

Algorithm for 2  n games:

If the given pay off matrix is 2  n (no. of rows are 2 and no. of columns are more than 2). Then
we follow the steps

Step 1: Plot the gain function on a graph assuming a suitable scale. Keep x on X-axis and gain on
y-axis.

Step 2: Since the player A is a maximum player, find the highest intersection point in lower
boundary of the graph. Let it be maximum point

Step 3: if the number of lines passing through the maximum point is only two, from 2  2 pay
off matrix form the original problem by retaining and go to step 8 otherwise go to step 7.

Step 4: identify any two lines with opposite slopes passing through that point. Then from 2  2
pay off matrix form the original problem by retaining only column corresponding to those two
lines which is having opposite slopes.

4
Step 5: Solve 2  2 game using oddments and find strategies for player ‘A’ and ‘B’ and also value
of game.

Algorithm for m 2 game:

If the given pay off matrix is m 2 (no. of rows are more than 2 and no. of columns are 2). Then
we follow the steps

Step 1: plot the gain function on a graph function on a graph by assuming a suitable scale, keep
‘y’ on ‘x-axis’ and gain on ‘y – axis’.

Step 2: since B is minimax player, find lowest intersection point in the upper boundary of the
graph. Let it be the minimax point.

Step 3: if number of lines passing through minimax point is only two, form a m 2 pay off
matrix form original problem by retaining only the row corresponding to those two lines and go
to step 8, otherwise go to step 7.

Step 4: identify any two lines with opposite slopes passing through that point the form 2  2
pay off matrix form the original by retaining only the rows corresponding to two lines which are

Step 5: solve the 2  2 game using oddments and find strategies for player ‘A’ and player ‘B’
and also the value of the game.

Dominance Method:

Some times it is observed that one of pure strategies of either player is always inferior
to at least one of remaining one. The superior strategies are said to dominate the inferior one.
Clearly a player would have to incentive to use inferior strategies which are dominated by
superior one. In such cases of dominance, we can reduce the size of pay off matrix by deleting
those strategies which are dominated by others. Thus if each element in one row say k th of pay
off matrix (aij) is less than or equal to the corresponding elements in some other row say r th
then player A will never choose k th strategy in other words probability Pk =P (choosing the kth
strategy) is zero, if akj < arj for all j =1,2,3,….,n

The value of the game and the non – zero choice of probabilities remain unchanged
even after the deletion of kth row from the pay off matrix. In such a case the kth strategy is said
to be dominated by r th one.

General rules of dominance are:

1) If all the elements of a row say rth less than or equal to the corresponding elements of
any other row say rth then kth is dominated by r the row.

5
2) If all elements of a column, say kth are greater than or equal to the corresponding
elements of any other column say rth then kth column is dominated by rth column.
3) Dominated rows and columns can be deleted to reduce the size of pay off matrix is
optimal strategies will remain unaffected.

The modified Dominance Property:

The dominance property is not always based on superiority of pure strategies only. A
given strategy can also be said to be dominated if it is inferior to an average of two or more
other pure strategies more generally. If some convex linear combination of some rows
dominates the ith row, then ith row will be deleted similar arguments follow for columns.

6
 Queuing Theory(Waiting Lines)

Queue called the waiting line in order to maintain a proper discipline. Here the arriving
people are called customers and the persons issuing the ticket are called a Server.

Queuing System:

A Queuing system can be completely described by

I. The input (or arrival pattern)

II. The service mechanism (or service pattern)
III. The Queue discipline
IV. Customer’s behavior.

The input pattern (or arrival pattern)

The input describes the way in which the customers arrive and join the system.
Generally the customers arrive in more or less random fashion which is not worth making the
prediction. Thus the arrival pattern can be described in terms of probability distribution for
inter arrival times (the time between two successive arrivals) must be defined. We deal with
those queuing system in which the customers arrive in ‘Poisson’ fashion. Mean arrival rate is
denoted by λ.

The service pattern (Out pattern)

It described how many customer can be served at a time, what the statistical
distribution of the service time is , and when the service is available. Service time may be a
constant or a random variable. Distributions of service time which are important in practice are
the exponential distribution. The mean service rate is denoted by µ.

The Queue Discipline:

The Queue discipline is the route determining the formation of the queue, the manner
of the customer’s behaviors while waiting, and the manner in which they are chosen for
service. The simplest discipline is “First come, first served”. Such type of Queue discipline is
observed at a ration shop.

Some of the Queue service disciplines are

FIFO – “First in, First out”

LIFO – “Last in, First out”

7
SIRO – “Service in Random order”

Customer’s Behavior:

The customer generally behaves in 4 ways

1. Balking: A customer may leave the Queue, of there is no waiting space

2. Reneging: This occurs when the waiting customer leaves the Queue due to Impatience.
3. Priorities: In certain applications some customers are served before others regardless of
their order of arrival.
4. Jockeying: Customers may jockey from one waiting line to another.

Queuing Models:

Generally Queuing Model may be completely specified in the following symbol form: (a/b/c:
d/e)

Where ‘a’ = inter arrival time

‘b’ = inter service time

‘c’ = Number of service channels

‘d’ = system capacity

‘e’ = Queue discipline

Model I M / M /1:  / FIFO

This model deals with a queuing system having single service channel. Poisson input,
Exponential service and there is no limit on the system capacity while the customers are served
on a “First in, First out” basis.

Characteristics of Model I

1. Average number of customers in system is given by


Ls =
 −
2. Average number of customer in Queue

2
Ls =
 ( −  )
8
3. Average waiting time in system

1
ws =
( −  )
4. Average waiting time in queue


wq =
 ( −  )

5. Probability of Queue size being greater than or equal to N

N

P (Queue size  N) =  

6. Average length of non – empty queue is given by

 m  
E =
 m  0   − 
7. Probability of ‘n’ customer in the system

n
    
Pn =   1 − 
    

8. Probability of no customers in the system

 
P0 = 1 − 
 
9. Expected waiting time of a customer who has to wait

1
E (W / W  0) =
 −
10. Probability ( waiting time in the system  t )

9
 
=  (  −  )e − (  − ) w dw
t

11. Probability ( waiting time in the Queue  t )

  
=    (  −  )e − (  − ) w dw
t


Model II M / M /1: N / FIFO

This model deals with a queuing system having single service channel. Poisson input,
Exponential service and there is ‘N’ customers in the system capacity while the customers are
served on a “First in, First out” basis.

Characteristics of Model II

1. Probability of ‘n’ customer in the system

1 − (  ) 
Pn =  ()
n

1 − (  ) 
N +1
 

Where  =


2. Probability of no customers in the system

P0 =
1 −  
1 −  N +1 

3. Average no. of customers in the system is

N
Ls = P0  n n
n=0

1− 
where P0 =
1 −  N +1
4. Average no. of customer in the Queue

10
 
Lq = Ls −

5. Average waiting time in the system

Ls
Ws =

6. Average waiting in the Queue
Lq
Wq =


Model III M / M / C :  / FIFO

This model deals with a queuing system having more than one service channels. Poisson
input, Exponential service and there is no limit on the system capacity while the customers are
served on a “First in, First out” basis.

Model IV M / M / C : N / FIFO

This model deals with a queuing system having more than one service channels. Poisson input,
Exponential service and there is ‘N’ customers in the system capacity while the customers are
served on a “First in, First out” basis.

Simulation
Simulation is a imitation of reality. It representation of reality through the use of a model or
other device which will react in the same manner as reality under a set of condition.

Examples:

a. Testing of medicines on animals in laboratories. In this case, the animal responses

simulate the human responses.

11
 b. Driving automobiles on test tracks. The test track simulates the actual environment in
which the automobiles are driven.

Monte-Carlo Technique or Monte – Carlo simulation

This technique involves the selection of random observations within the simulation
model. It is constrained for application involving random numbers to solve deterministic and
stochastic problems the underlying principle of this technique is

1) Replace the actual statistical universe by another universe described by some assumed
probability distribution.
2) Sample from this theoretical population by mean of random numbers.

Steps in Simulation

Step 1: Clearly define the problem

a) Identify the objectives of the problem.

b) Identify the main factors which have the greatest effect on the objectives of the
problem.

Step 2: Construct an appropriate model.

a) Specify the variable and parameters of the model

b) State the conditions under which the experiment is to be performed.
c) Define the relationship between the variables and parameters.

Step 3: Prepare the model for experimentation

a) Define the starting conditions for the simulation.

b) Specify the number of runs of simulation to be made.

Step 4: Using step 1 to 3, experiment with the model.

a) Define a coding system that will correlate the factors defined in step with the random
numbers to be generated for the simulation
b) Select a random number generator and create the random numbers to used in the
simulation.
c) Associate the generated random numbers with the factors identified in step 1 and
coded in step 4 (a).

Step 5: Summarizes and examine the results obtained in step 4.

12
 Step 6: Evaluate the results of the simulation

Advantages to Simulation:

• Can be used to study existing systems without disrupting the ongoing operations.

• Proposed systems can be “tested” before committing resources.

• Allows us to control time.

• Allows us to identify bottlenecks.

• Allows us to gain insight into which variables are most important to system
performance.

• Manufacturing: Material handling systems, assembly lines, automated production

facilities, inventory control systems, plant layout, etc..

• Business: Stock and commodity analysis, pricing policies, marketing strategies, cash
flow analysis, forecasting, etc..

• Government: Military weapons and their use, military tactics, population forecasting,
land use, health care delivery, fire protection, criminal justice, traffic control, etc.

Disadvantages to Simulation

• Model building is an art as well as a science. The quality of the analysis depends on the
quality of the model and the skill of the modeler (Remember: GIGO)

• Simulation results are sometimes hard to interpret.

•
Simulation analysis can be time consuming and expensive should not be used when an
analytical method would provide for quicker results.

13
UNIT-5
 UNIT-6
Dynamic Programming
Introduction: Dynamic programming is to perform optimization (Minimization or
Maximization). It is to divide the original problem into many sub problems and solve the sub
problems individual obtain the optimum solution of the last problem by integrating the
optimum solution of original problem.

Characteristic of DPP

The features which characterize the dynamic programming problem are as follows:

1. The problem can be divided into stages with a policy decision required at each stage.
2. Every stage consists of a number of states associated with it. The states are the different
possible condition in which the system may find itself at that stage of the problem
3. Decision at each stage converts the current stage into state associated with the next
stage.
4. The state of the system at a stage is described by a set of variables called state variables.
5. When the current state is known, an optimal policy for the remaining stages is
independent of the policy of the previous ones.
6. The solution procedure begins by finding the optimal policy for each state to the last
stage.
7. A recursive relationship which identifies the optimal policy for each state with n stages
remaining, given the optimal policy for each state with (n-1) stages left
8. Using recursive equation approach, each time the solution procedure moves backward
stage by stage by stage for obtaining the optimum policy of each state for the particular
stage, till it attains the optimum policy beginning at the initial stage.

Dynamic Programming Algorithm

The computational procedure for solving a problem by dynamic programming approach can be
summarized in the following steps:

Step 1: Identify the decision variables and specify objective function to be optimized under
certain limitations, if any

Step 2: Decompose (or divide) the given problem into a number of smaller sub-problems (or
stages). Identify the state variables at each stage and write down the transformation function
as a function of the state variables and decision variables at the next stage.
Step 3 : write down a general recursive relationship for computing the optimal policy decide
whether forward or Backward method is to follow to solve the problem

Step 4: Construct appropriate stages to show the required values of the return function.

Step 5: Determine the overall optimal policy or decisions and its value at each stage there may
be more than one such optimal policy.

Applications of Dynamic Programming:

➢ In the production area, this technique has been used for production scheduling and
employment smoothing problems
➢ It has been used to determine the inventory level and for formulating the inventory
recording
➢ It can be applied for allocating the scarce resources to different alternative uses such as,
allocating the salesmen to different sales districts etc.
➢ It is used to determine the optimal combination of advertising media(TV, Radio, News
papers) and the frequency of advertising.
➢ It can apply in Replacement theory to determine at which age the equipment is to
replace for optimal return from the facilities.
➢ Spare part level determination to guarantee high efficiency utilization of expensive
equipment.
➢ Other areas: Scheduling methods, Monrovian decision models, infinite stage system,
probabilistic decision problem etc.

Simulation
Simulation is a imitation of reality. It representation of reality through the use of a model or
other device which will react in the same manner as reality under a set of condition.

Examples:

a. Testing of medicines on animals in laboratories. In this case, the animal responses

simulate the human responses.
b. Driving automobiles on test tracks. The test track simulates the actual environment in
which the automobiles are driven.

Monte-Carlo Technique or Monte – Carlo simulation

 This technique involves the selection of random observations within the simulation
model. It is constrained for application involving random numbers to solve deterministic and
stochastic problems the underlying principle of this technique is

1) Replace the actual statistical universe by another universe described by some assumed
probability distribution.
2) Sample from this theoretical population by mean of random numbers.

Steps in Simulation

Step 1: Clearly define the problem

a) Identify the objectives of the problem.

b) Identify the main factors which have the greatest effect on the objectives of the
problem.

Step 2: Construct an appropriate model.

a) Specify the variable and parameters of the model

b) State the conditions under which the experiment is to be performed.
c) Define the relationship between the variables and parameters.

Step 3: Prepare the model for experimentation

a) Define the starting conditions for the simulation.

b) Specify the number of runs of simulation to be made.

Step 4: Using step 1 to 3, experiment with the model.

a) Define a coding system that will correlate the factors defined in step with the random
numbers to be generated for the simulation
b) Select a random number generator and create the random numbers to used in the
simulation.
c) Associate the generated random numbers with the factors identified in step 1 and
coded in step 4 (a).

Step 5: Summarizes and examine the results obtained in step 4.

Step 6: Evaluate the results of the simulation

Advantages to Simulation:
 • Can be used to study existing systems without disrupting the ongoing operations.

• Proposed systems can be “tested” before committing resources.

• Allows us to control time.

• Allows us to identify bottlenecks.

• Allows us to gain insight into which variables are most important to system
performance.

• Manufacturing: Material handling systems, assembly lines, automated production

facilities, inventory control systems, plant layout, etc..

• Business: Stock and commodity analysis, pricing policies, marketing strategies, cash
flow analysis, forecasting, etc..

• Government: Military weapons and their use, military tactics, population forecasting,
land use, health care delivery, fire protection, criminal justice, traffic control, etc.

Disadvantages to Simulation

• Model building is an art as well as a science. The quality of the analysis depends on the
quality of the model and the skill of the modeler (Remember: GIGO)

• Simulation results are sometimes hard to interpret.

•
Simulation analysis can be time consuming and expensive should not be used when an
analytical method would provide for quicker results.