Operations Research
Operations Research
Module 1
i. Physical Model (iconic model): -It includes all form of diagrams. Graphs and charts.
They are designed to deal with specific problems.
ii. Mathematical Model: - It is known as symbolic models also. It employs a set of
mathematical symbols to represent the decision variable of the system. They are
most widely used in OR. Mathematical model consists of a set of equations which
define and specify the relationship and interaction among various elements of
decision problem under study.
iii. Analogue model: - A set of properties are used. An organization chart is common
analogue model.
iv. Deterministic model: - Everything is defined and the results are certain. Certainty
is the state of nature assumed in these models.
v. Probabilistic model: - Input and output variables follow a probability distribution.
Operation Research Techniques
• Linear Programming Problem
• Transportation Problem
• Assignment problem
• Sequencing Problem
• Network problem
Limitations of OR
• OR tries to find out the optimal solution taking all the factors into account. In the era
of modern age, there are number of factors. Huge calculations are required to express
the relationship among them.
• Non-Consideration of intangible factors: - OR provides solution only if the elements
are qualified. Intangible factors cannot be quantified.
• Job of specialists: - Or is related with the job of specialists. But they are not
management experts.
• Costly: - There is a probability of having frequent changes in basic data. So, the
incorporation of these data in OR models become a costly affair.
• Difficult in implementation: - There will be resistance from employees.
• Difficult in selection of techniques
• Not a substitute for management
• Distance between operation researcher and manager: - OR being specialists job requires
a mathematician or a statistician, who might not be aware of the business problem.
Similarly, manager fails to understand complex working of OR.
• Dependance on an electronic computer: - Operations only be handled by computer.
Module 2
Linear Programming Problem
Linear programming (LP) or Linear Optimisation may be defined as the problem of
maximizing or minimizing a linear function that is subjected to linear constraints. The
constraints may be equalities or inequalities. The optimisation problems involve the calculation
of profit and loss.
1. Graphical method
2. Simplex method
Graphical Method
Graphical method is used to solve linear Programming Problem if it involves two variables.
Each constraint is represented by a line. If there are more than one constraint, many lines are
to be drawn.
Feasible Region
It is the area which satisfies all the constraints simultaneously. For >= constraints, the feasible
region will be the area which lies above the constraint line. If it is <= constraints, then area is
generally below these lines.
Simplex method
Simplex method is an interactive procedure. Interactive means step by step. We want to proceed
on a systematic manner from as initial basic feasible solution to another basic feasible solution,
till we reach to final solution. Following are the main steps
• Find the trail basic feasible solution
• Test whether it is feasible or not
• It is not feasible or optimal, then improve the first trail basic feasible solution by using
a set of rules.
• Repeat step 2 and 3 till an optimal solution is attained.
Basic Terms Used
1. Standard form: - All the constraints in the LPP is to be written as equalities.
2. Slack variable
If constraints have a sign <=, then in order to make it equality, we have to add a variable to the
left-hand side.
Eg: 4x1+2x2 <=40
4x1+2x2 +s1 = 40
3. Surplus variable
If constraints have a sign >=, then in order to make it equality, we have to subtract a variable
from the left-hand side.
E.g.: -: 4x1+2x2 >=40
4x1+2x2 - s1 = 40
4. Basic solution
A general LP problem with ‘n’ variable and ‘m’ constraints have a basic solution. It can be
found by setting ‘n-m’ of the variable equal to 0 and solving the constraints equations for the
values of the other ‘m’ variables. It is basic solution.
5. Basics
These are the variables whose values not restricted to 0 in the current basic solution. These are
listed in one column of the simplex table. The variable which makes up ‘basic’ are termed as
basic variable.
6. Zj Row
It is the row containing the figures for gross profit or loss given up by adding one unit of a
variable into the solution. Zj is the sum of the product of the coefficients of basic variables of
the objective function and the vector xj.
7. Vector
Any column or row of a simplex table is called a vector.
8. Net Evaluation or Index Row (∆j)
∆j is the net profit or loss if one unit of the variable in the respective column is introduced. It is
also known as cj-zj,
∆j = cj-zj
9. Pivot or key column or Incoming vector
The column has highest positive ∆j in maximization problem and highest negative ∆j in
minimization problem. It is called incoming vector.
10. Minimum Ratio
It is the lowest non negative ratio in the replacing ratio column. Replacing ratio column
contains values obtained by dividing each element in xB column.
11. Pivot or Key row
It is the row relates to the minimum ratio.
12. Key element
It is that element of simplex table which lies both in the key row and key column.
Artificial variable
Artificial variables are fictitious variables which are introduced only for computation purpose.
It has no physical meaning or economic significance.
Eg:- In a minimization problem the constraints are
5x1 + 4x2 + 2x3 >= 25
2x1 + 3x2 + x3 =32
Rewrite the above constraints as follows
5x1 + 4x2 + 2x3 – s1 + A1 = 25
2x1 + 3x2 + x3 + A2 =32
Unbounded solution
If we notice that every variable in the incoming variable column has all its elements negative
or zero when we are searching the outgoing variable, we are face to face with a problem which
have unbounded solution.
Infeasibility solution
If a LPP has no optimal solution, we can say that infeasibility exists. In final solution an
artificial variable exists in basis as a positive value, then there is no feasible solution.
Duality Method
Duality describes that LP problem exist in pairs. So every LPP has another LPP which is related
to it. This related LPP is called dual LPP and original LPP is called Primal LPP. As per duality
theorem for every maximization(minimization) problem in linear programming, there is a
unique similar problem of minimization (maximization) involving the same data which
describes the original problem. The optimal solution from the both primal and dual would be
the same because they originate from the same data.
Module 3
Transportation Model
Transportation problem deals with transportation of products manufactured at different
factories or plants to a number of demand destinations. The main objective is to satisfy the
demand destination requirements with in the plant capacity constraints at minimum
transportation cost.
Methods to find out initial feasible solution
There are three method to find out initial basic feasible solution
1. The North-West corner method
2. Least cost method
3. Vogel’s Approximation method
North-West Corner Method
It is the simplest method, which is used to find generate an initial feasible solution.
Step 1:- Consider the plant supply or availability and first market or warehouse demand. The
lowest among two are assigned to the first plant, first market or demand route. If supply is more
than demand, shift to the next market and if supply is less than demand, then move to the next
plant. If supply and demand is equal, move diagonally to the next plant and the next market
route. So either a column or row total is to be exhausted.
Step 2:- Consider the reduced matrix. Then continue the steps explained in ‘1’ till we get an
initial feasible solution to the problem.
Least-Cost Method or Matrix Minima method
In this method allocations are made in the cells with the smallest unit costs. It reduces the
computation as well as the amount of time necessary to arrive at the optimal solution.
Step 1:- Select the cell with smallest transportation cost among all the rows and columns of the
transportation table
Step 2:- If there are more than two lowest same cost, we can select the cells for allocation
arbitrarily, among the lowest cost cells.
Step 3:- Assign maximum units in the lowest cost cell. Then we can axhaust either row or
column on the basis of allocation. Eliminate that column or row.
Step 4:- Consider the reduced matrix table and select another lowest cost cell. The allocate tha
maximum units in that cell. On the basis of that we can exhaust either a row or column.
Step 5:- Consider the process till all available quantities are exhausted.
Vogel’s Approximation Method
It is based on the concept of opportunity cost.
Step 1:- Find the penalty cost namely the difference between the smallest and next smallest
cost in each row and column.
Step 2:- Among the penalties found in (1) select the maximum penalty . If there is a tie relate
with maximum penalties, select any one arbitrarily.
Step 3:- From the selected row or column, find out the cell which is having lowest cost. As
much as quantity allocated to this cell by considering the demand and supply.
Step 4:- The column or row which is exhausted is to be deleted.
Above steps repeated till initial solution is attained.
Degeneracy in Transportation problem
Some times the number of non-negative independent allocation is less than m+n-1, in
transportation problem. It may happen either at the initial stage or at subsequent iteration. It is
called degeneracy.
To resolve degeneracy, the following steps are to be adopted.
1. Among the empty cells, choose one empty cell having at least cost, which is of an
independent position. If these cells are more than one, it is better to select any of them,
arbitrarily.
2. A small positive quantity is to be allocated in the selected cell.
3. The cell containing the above small positive quantity will be treated like other occupied
cells and degeneracy is to be removed by adding one or more accordingly.
Transshipment Model
Transshipment problem, which allows for the shipment of goods both from one source to
another and from one destination point to another. For example, a multi-plant firm may find it
necessary to send some goods from one plant to another in order to meet the substantial increase
in the demand in the second market. The second plant here would act both as a source and a
destination and there is no real distinction between source and destination. Sometimes a point
in the shipment process can both receive goods from other points and send goods to other
points. This point is called transshipment point through which goods can be transshipped on
their journey from a supply point to demand point. The optimal solution to a transshipment
problem can be found by converting this transshipment problem to a transportation problem
and solve.
Module 4
Assignment model
Assignment problem is a specialtype of linear programming problem. The goal of the
assignment problem is to minimize the cost or time of completing a number of jobs by a number
of persons.
1. Only one job is assigned to person.
2. Each person is assigned with exactly one job.
Assumptions
1. The number of assignees and the number of tasks or jobs are the same.
2. Each assignee is to be assigned to exactly one task.
3. Each task is to be performed by exactly exactly one assignee.
4. There is a cost Cij associated with assignee performing task.
5. The objective is to determine how all n assignments should be made to minimize the
total cost.
Hungarian method
Hungarian method of assignment provides an efficient means of finding the optimal solution
without having to make a direct comparison of every option. This method based on the concept
of opportunity cost. As per the Hungarian method, the following steps are to be followed to
solve an assignment problem
1. Develop the opportunity table
a. Row Reduction: Subtract the minimum entry of each row from all the entries of
the respective row in the cost matrix.
b. Column Reduction: After row reduction, subtract the minimum entry of each of
the column from all the entries of the respective column.
2. Draw the minimum number of horizontal and vertical lines that are required to cover
all zero elements. The solution is optimal if the number of lines drawn is equal to n
(number of row / column) and proceed to step number 5. If the number of lines is less
than n, then proceed to step 3.
3. Select the smallest cost of uncovered lines. Subtract this element from all uncovered
elements including itself and add this element to each value which is located at the
intersection of any two lines. The cost element through which one-line passes remain
unaltered.
4. Repeat step number 2 and 3 until an optimal solution is obtained.
5. Make the job assignment as indicated by the zero elements in the given optimal
solution. For this purpose
a. Locate the row which contain only one zero element. Assign the job
corresponding to this element to its corresponding person by . Cross out the
zeros if any the column corresponding to the element. When zeros are cross out,
it is an indication that the particular job and persons are no more available.
b. Repeat the above step for each of such rows which contain only one zero.
Similarly perform the same operation. Likewise perform the same operation in
respect of each column containing only one zero element and cross out the other
zeros of the corresponding rows.
c. If there is no row or column with only a single zero element, then select
row/column arbitrarily and choose of the jobs and make the assignment
d. Calculate the total cost with reference to original table.
Maximization case in assignment problem
Assignment problem may use for maximization of profit or revenue instead of costs as the
objective. To deal with maximization problem, first convert the problem into a minimization
problem. For this purpose, subtract each element of entries from the highest value of the table.
Module 5
Network analysis
A project is composed of jobs, activities or functions related to one another and all those should
be completed in order to complete a project. It is a popular technique used for planning,
scheduling, monitoring and coordinating large and complex projects comprising a number of
activities.
Objectives of network analysis
1. It helps to minimize time of the completion of the project
2. It helps to minimize time for a given cost
3. It helps to minimize cost for a given total time
4. It avoids production delays, interruptions and conflicts
5. It leads to optimal use of resources
Network Techniques
Network technique were developed to facilitate planning, scheduling, the project in an
integrated manner so that those should be completed with in the constraints of desired time,
cost and performance. Two most popular techniques are CPM (Critical Path Method) and PERT
(Project Evaluation and Review Techniques).
Basic components
Network: Network is a series of related activities which result in some product or services.
Network diagram: It is a diagram which the activities and events of a network, their sequence
and relationship. It shows the activities and events of a project in a local sequence.
Activity: Activities of the network represents project operation or tasks to be conducted. Each
activity is represented by an arrow( ).The head of the arrow indicates the sequence or flow
of activities. Activities can be classified into three categories
i) Predecessor Activity: An activity must be completed before one or more other
activities start.
ii) Successor activity: Activities that cannot be started until one or more of the
other activities are completed.
iii) Dummy Activity: Any activity which does not consume either any resource or
time is called dummy activity. It is needed when (i) two or more parallel
activities in a project have same head and tail events (ii) two or more activities
have the common immediate predecessor activities. Dummy activities
represented by dotted arrows.
Events: An event represents the start and end of an activity. Events commonly represented by
circles. The event can be classified into two categories
i) Merge Event: An event which represents the joint completion of more than one
activity
ii) Burst Event: An event which represents the initiation or beginning of more than
on activity.
Critical Path Method
CPM networks are activity-oriented diagrams. It is analyzed to locate the critical path. The
main objective of critical path analysis is to estimate the total project duration and to assign
starting and finishing times to all activities involved in the project
Basic terms
Advantages
1. It helps top management to concentrate their attention to the critical activities and their
completion in time.
2. It helps to locate the critical and non-critical activities.
3. It helps to plan and schedule a project.
4. It provides relevant information of an activity.
Limitations
1. Time of each activity is known. But in real practice it is not true
2. It does not consider the statistical analysis.
3. It is suitable in a project where there is a definite start and definite finish.
9. Sum of the variance of time estimates of all activities are the critical activities.
10. Calculate the standard normal variable Z
Sequencing Model
It is used to obtain the most feasible solution by performing the series of jobs on different
available machines. It is concerned with a suitable selection of a sequence of jobs to be done
on a finite number of service facilities in some well-defined technological order so as to
optimize some efficiency measure such as total elapsed time or overall cost.
Assumptions
1. The processing time on different machines are independent of the order of the job in
which they are to be processed.
2. Only one job can be processed on a given machine at a time.
3. The time taken by the job in moving from one machine to another is very negligible
and is taken as equal to zero.
4. Each job once started on a machine is to be performed up to the completion on that
machine.
5. Machine to be used are different types
6. Processing times are given, do not change
7. The order of completion jobs is no significance
Basic Terminology
1. Processing time means the time each job requires at each machine
2. Total elapsed time is the time between starting the first job and completing last job
3. Idle time on a machine is the time a machine remains idle during the total elapsed time.
4. No passing rule implies that passing is not allowed ie, the same order of jobs is
maintained over each machine.