PG - or - 1 - LPP - 1 - 24
PG - or - 1 - LPP - 1 - 24
PG - or - 1 - LPP - 1 - 24
OPERATION RESEARCH
Unit-V - Operations Research - Syllabus
Linear programming - Simplex Computational procedure - Geometric interpretation
of the simplex procedure - The revised simplex method - Duality problems - Degeneracy
procedure - Peturbation techniques - integer programming - Transportation problem – Non-
linear programming – The convex programming problem - Dynamic programming -
Approximation in function space, successive approximations - Game theory - The maximum
and minimum principle - Fundamental theory of games - queuing theory / single server and
multi server models (M/G/I), (G/M/I), (G/G1/I) models, Erlang service distributions cost
Model and optimization - Mathematical theory of inventory control - Feedback control in
inventory management - Optional inventory policies in deterministic models - Storage
models – Dam type models - Dams with discrete input and continuous output - Replacement
theory – Deterministic Stochostic cases - Models for unbounded horizons and uncertain case
- Markovian decision models in replacement theory - Reliability - Failure rates - System
reliability - Reliability of growth models - Net work analysis - Directed net work - Max flow
min cut theorem - CPM-PERT - Probabilistic condition and decisional network analysis.
Introduction
The germination of the concept Operations Research occurred during World War I.
In England in the year 1915 F.W.Lanchester attempted to treat military operations
quantitatively. He derived equations relating the outcome of a battle to both the relative
numerical strength of the combatants and their relative manpower. He modeled a
situation involving strategic choices and then tested that model against a known
situation.
During the same period, Thomas Alva Edison in America was studying the process of
Anti-submarine warfare. He devised a war game to be used for simulating problems of
naval Manoeuvre.
In 1915 F.W. Harris had developed the first model on an inventory problem for economic
lot size.
In 1917 A.k.Erlang, a Danish Mathematician has developed solutions for some waiting
line problems.
In 1930 W.Leontieff developed a linear programming model representing the entire
United States economy.
Active research works were done during World War II in Great Britain and United
States of America. During World War II, military management called on scientists from
various disciplines and organized them into teams to discuss and suggest ways and
means to improve the execution of various military projects.
Following the end of the word war, the success of military team attracted the attention of
industrial managers who were seeking solutions to their complex executive type
problems.
The term Operations Research (O.R) was first found by Mc Closky and Trefthen in 1940.
In 1947 George B.Dantizing developed simplex solution to linear programming problem,
earlier formulated by Leontieff.
During the year 1950, O.R was first recognized as a subject in universities.
In 1958 network models were developed by the U.S special project office.
Some Definitions of O.R
D’Clarke’s - “The art of Winning war without actually fighting”
T.L.Sastry - The art of giving bad answers to problems where otherwise worse answers
are given.
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 1
Operation Research Linear Programming Problem
Introduction
Definition of LPP
Linear programming problems are the problems that deal with maximization or
minimization of a numerical function of a numbers of variables subject to a set of linear
equations.
Note: The objective function, set of constraints and the non-negative constraint together
form a linear programming problem.
Maximize or Minimize
Subject to ,
and for
Maximize or Minimize
Subject to ,
Where , B=
Characteristics of LPP
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 2
Operation Research Linear Programming Problem
Graphical Method
A linear programming problem with only two decision variables can be easily solves
by graphical method.
1. Draw one decision variable in axis and another decision variable in axis with
appropriate scale
2. Consider each inequality constraint as equality equation
3. Plot each equation on the graph as each will geometrically represent a straight line.
Every point on the line will satisfy the equation of the line.
4. If the inequality constraint is , then shade the region below the line. If the inequality
constraint is , then shade the region above the line in the first quadrant shade the
common region which will satisfy all the constraints simultaneously. The common
region thus obtained is called the feasible solution.
5. Choose the value and plot the objective function line, which is called Iso-profit
line
6. In the maximization type problems, the Iso-profit line parallel to the objective
function line farthest from the origin and passing through at least one corner of the
feasible region. In the minimization type the shifting line will stop nearest to the
origin and passing through at least one corner of the feasible region.
7. Find the co-ordinates of the extreme points selected in step 6 and substitute this in
the objective function equation to get the maximum or minimum.
If a linear programming problem has more than one optimal solution, it is said to
have multiple or alternative optimal solutions.
c) Unbounded Solution
d) No feasible Solution
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 3
Operation Research Linear Programming Problem
PG-TRB- Questions
2. Graphical Method is usually used to find the solution of a linear programming problem
if the objective function contains [2004-05]
a) two variables b) three variables
c) four variables d) only one variable
; is :
a) Bounded solution b) Unbounded solution
c) Infeasible solution d) Optimal solution
4. In a graphical solution of LPP, the optimal is at atleast one of the points [2015]
a) on the boundary of the common region b) in the common region
c) at the corner points of the boundary d) none of these
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 4
Operation Research Linear Programming Problem
Maximize Z=
Subject to ,
and for
Minimize
Subject to ,
Where , B=
Any LPP can be put in the canonical form by the use of some elementary transformations.
For Example
is equivalent to
with
Note: All LPP the objective function can be expressed in the maximization form.
For Example
The constraints is equivalent to
Also is equivalent to
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 5
Operation Research Linear Programming Problem
Example1:
Maximize
Subject to ; ; and
Canonical Form
Maximize
Subject to ; ; and
Example2:
Minimize
Subject to ; ; and
Canonical Form
Maximize
Subject to ; ; and
Example3:
Maximize
Subject to ; ;
And ;
Canonical Form
Maximize
Subject to ; ;
and
PG-TRB- Questions
a) Maximize
Subject to ; ;
b) Maximize
Subject to ; ;
c) Minimize
Subject to ; ;
d) Maximize
Subject to ; ;
a) b)
c) d)
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 6
Operation Research Linear Programming Problem
Subject to ,
and for
Minimize
Subject to ,
Where , B=
The variable which are added to the left hand side of the inequality constrains
involving less than or equal to sign , are called slack variables.
; where
Surplus variable
The variable which are subtracted from the left hand side of the inequality constrains
involving greater than or equal to sign , are called slack variables.
; where
Example1:
Maximize
Subject to ; ; and
Standard Form
Maximize
Subject to ; ;
And ;
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 7
Operation Research Linear Programming Problem
Example2:
Maximize
Subject to ; ;3
And ;
Standard Form
Maximize
Sub to ; ;3
And ; ;
Example3:
Miximize
Subject to ; ;
And ;
Standard Form
Minimize
Subject to ; ;
and ;
PG –TRB – Questions
3. The total number of slack or surplus variables to be introduced to write the following
linear programming problem [2005-06]
Maximize
Subject to
In standard form is
a) 1 b) 2 c) 4 d) 3
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 8
Operation Research Linear Programming Problem
Theorem
A set of points is said to be a convex set if for any two points in the set , the line
segment joining them is also in the set.
Thus a set is a convex set if convex combination of any two points in is also in
i.e,
Basic Solution
The (any of them may be zero) variables are called basic variable and zero
variables are called non-basic variables.
Any basic feasible solution which optimizes the objective function (maximize or
minimize) is called its optimal solution.
PG –TRB – Questions
1. Any solution of Linear programming problem which satisfies the non-negative
restrictions is called: [2012-13]
a) Optimal solution b) Non-feasible solution
c) Feasible solution d) Basic solution
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 9
Operation Research Linear Programming Problem
Simplex Method
Introduction
When the numbers of decision variables are more than two, graphical method
becomes ineffective. The mathematical technique in LPP called the Simplex method was
developed in 1947 by an American Mathematician, G.B.Dantzig. This method is simplex and
widely used for solving a problem which has two or more variables.
PG-TRB- Questions
2. The number of basic feasible solutions of three equations in four unknowns is [2015]
a) 12 b) 4 c) 7 d) 6
a) b) c) d)
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 10
Operation Research Linear Programming Problem
Example - 1
Maximize
Subject to, ;
Find the following
i) Basic feasible solution
ii) Is the solution feasible and optimal?
iii) Optimal basic feasible solution
iv) Is the non-degenerate solution feasible?
Solution
Here and , then the basic solutions are obtained by setting
variable equal to zero and the remaining two variables are solved.
The number of possible basic solutions is are given by
Values of
Values of the basic Is the Is the Is the
Sl. Basic Non- Basic the
variables from the solution solution non- solution
No. Variables Variable objective
constraint equations feasible? degenerate? optimal?
function
2 No Yes No
,
3 2 Yes Yes No
,
Example - 2
Maximize
Subject to,
3 2 No Yes No
,
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 11
Operation Research Linear Programming Problem
Example 3:
Solution
Here, ; ; and
When
When
When
Example 4:
Solution
Here, ; ; and
When
When
When
Example 5:
Solution
It is obvious that the objective function for the basic feasible solution is given by
.
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 12
Operation Research Linear Programming Problem
Example 6:
Let , be a feasible solution to the system of equations
and reduce the given feasible solution to a basic
feasible solution
Solution
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 13
Operation Research Linear Programming Problem
Simplex Algorithm
For the solution of any LPP by simplex algorithm, the existence of an initial basic
feasible solution is always assumed. The step for the computation of an optimum solution is
as follows.
… 1 0 0
… 0 1 0
… 0 0 1
. . . . . . . . .
. . . . . . . . .
Here, Column 4,5,6 – Body matrix Column 7,8,9 - Unit matrix
Where
Row denotes the cost coefficients of the variables in the objective function.
Column denotes the cost coefficient of the basic variables in the objective function.
column denotes the basic variables
column denotes the values of the basic variables
Let denotes the coefficients of the non basic variables in
the constraint equations give the body matrix.
The coefficients of the basic variables give the unit matrix which is shown in
the above table.
5. Find the net evaluations where by using the relation
and examine the sign of this difference.
i) If all then the current basic feasible solution is optimal
ii) If at least one then the current solution is not optimal
Go to the next step.
6. Choose the most negative net evaluation . If tie exists, select arbitrarily any one.
The column which the most negative value is known as Key column or pivot
column.
7. Compute the minimum ratio (ie) . It is the ratio between the solution column
number and the key column number.
i) If then there will be an unbounded solution to the LPP. If ,
select this row which is known as key row or pivot row. Here, the key
column variable is entering variable and the key row variable is going to
leave from the table is known as leaving variable. The intersecting of key row
and key column is known as key number.
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 14
Operation Research Linear Programming Problem
The element at the intersection of the key row and key column is called
pivot or key element.
8. Eliminate the leaving variable and introduce the entering variable in the table
along with its associated value under column. Find the next table values as
shown below.
i) New key row number
ii) Other than key row numbers
New number –
9. Go to step 5 and repeat the procedure until to get an optimal solution or there is
an indication of an unbalanced solution.
Maximization:
Minimization:
PG-TRB- Questions
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 15
Operation Research Linear Programming Problem
Artificial Variable
So far we have solved problems which have slack variables readily provided the
initial basic feasible solution? There are however, many LPP where slack variables cannot
provide the initial basic feasible solution. In these problems, at least one of the Constraints if
of the type or = signs after ensuring all bi .
To obtain an initial basic feasible solution any non negative variable which is
introduced in the constraint in order to get the initial basic feasible solution is called artificial
variable.
Note:
1. We need artificial variables for to get the initial basic feasible solutions.
2. Negative values for slack variables are unacceptable.
Example-1
Maximize
Subject to ; ; ;
To form the initial simplex table the number of artificial slack variables to be added is
Solution
Subject to
To form the initial simplex table the number of artificial slack variables to be added is
a) 3 b) 2 c) 1 d) 0
2. In the following linear programming problem [2004-05]
Minimize
Subject to
The number of artificial variable(s) used to solve the linear programming problem by
simplex method is
a) 1 b) 2 c) 3 d) 0
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 16
Operation Research Linear Programming Problem
1. Express the linear programming problem in the standard form by introducing slack
and / or surplus variables, if necessary.
2. To obtain an initial basic feasible solution, add nonnegative variables to the left hand
side of all the constraints of type. These variables are called artificial
variables. To achieve the final solution, these artificial variables are assigned a large
penalty of –M for maximization and +M for minimization problems in the objective
function.
3. Solve the modified linear programming problem by simplex procedure following
three cases may arise while making iterations.
i) If no artificial variable appears in the basis and the optimality condition
satisfied, then the current solution is an optimal basic feasible solution.
ii) If at least one artificial variable appears in the basis at zero level and
optimality condition is satisfied, then the current solution is an optimal basic
feasible (though degenerated) solution.
iii) If at least one artificial variable appears in the basis at non zero level and
optimality condition is satisfied, then the problem has no feasible solution.
The solution satisfies the constraints but does not optimize the objective
function, because a large penalty M is assigned to this artificial variable.
Hence it is called pseudo – optimal solution.
Example-1
Maxi
Subject to ; ; ; to
form the initial simplex table the number of artificial slack variables to be added is 2
Solution
Maxi
Subject to ;
Example-2
Mini
Subject to ; ; ; to form the
initial simplex table the number of artificial slack variables to be added is 2
Solution
Mini
Subject to ;
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 17
Operation Research Linear Programming Problem
Phase-I
In this phase, construct an auxiliary LPP leading to a final simplex table containing a
basic feasible solution to the original problem.
Step1: Assign a cost ‘-1’ to each artificial variable and a cost ‘0’ to all other variables and get
a new objective function.
Step2: Construct the auxiliary LPP in which the new objective function is to maximized
subject to the given set of constraints.
Step3: Solve the auxiliary LPP by simplex method. The following three cases arise
i) Max and at least one artificial variable appear in the optimum basis at
positive level. In this case the given LPP does not possess any feasible
solution, stop the iterative procedure.
ii) Max and at least one artificial variable appear in the optimum basis at
zero level. In this case proceed to Phase II.
iii) Max and no artificial variable appear in the optimum basis. In this case
also proceed to Phase II.
Phase II:
Use the optimum basis feasible solution of Phase - I as an initial solution for the
given LPP. Assign the actual costs to the variables in the objective function and zero cost to
every artificial variable that appears in the basis at zero level. This modified simplex table is
solved by simplex procedure until an optimum basis feasible solution (if exists) has been
obtained. The new objective function must be of maximization type. Remove all artificial
variables from the table at the end of phase-I before starting phase –II.
Example-1
Maxi
Subject to ; ; ;
By using Two Phase method, to form the initial simplex table the number of artificial
slack variables to be added is
Solution
Maxi
Subject to ;
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 18
Operation Research Linear Programming Problem
Step2: Compute an initial basic feasible solution with an initial basis and form the
auxiliary matrix such that and
Step3: Consider the object function as an additional constraint and form such
that and
Step5: Compute
Step6: Write down the results obtained from step 2 to step 5 in the revised simplex table.
Step7: Convert the leading element to unity and the remaining all other elements of the
entering column to zero. Find the current improved feasible solution.
Step8: Repeat the same procedure until an optimum basic feasible solution is obtained or to
get an unbounded solution.
PG-TRB- Questions
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 19
Operation Research Linear Programming Problem
Duality Theorems
1. The dual of the dual is Primal
2. For every primal problem there is a dual problem.
3. If dual has no feasible solution, then primal also admits no feasible solution.
4. If either of the primal or the dual problems has an unbounded solution then the
solution to the other problem is infeasible or no solution.
5. The necessary and sufficient condition for any LPP and its dual to have optimum
solution is that both have feasible solutions. Moreover if one of them has a finite
optimum solution, the other also has a finite optimum solution.
6. If both the primal and the dual problems have feasible solution, then both have
optimal solutions and Max Z = Min Z’. ie.
7.
8.
9. Complementary slackness theorem.
i) If a primal variable is positive, then the corresponding dual constraint is an
equation at the optimum and vice versa.
ii) If the primal constraint is a strictly inequality, then the corresponding dual
variable is zero at the optimum and vice versa.
10. The values of the objective function in both the problems are same. ie Max Z = Min Z’.
The solution for the variables of the other problem can be read from the
row below the columns of slack and surplus variables.
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 20
Operation Research Linear Programming Problem
Primal Dual
Nature of Nature of
Objective Objective
Constraints Constraints
function function
decision variable decision variable
Maximize Minimize
Subject to Subject to
Maximize Minimize
Subject to , Subject to ,
Where ; ; B= ;
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 21
Operation Research Linear Programming Problem
Example-1
Maximize
Subject to ; ; ; .
Solution
Minimize
Subject to ; ;
Example-2
Minimize
Subject to ; ; .
Solution
Maximize
Subject to ; ; ;
Example-3
Minimize
Subject to ; ; ; .
Solution
Maximize
Subject to ; ;
PG – TRB – Questions
1. The dual of the following linear programming problem [2002-03] & [2011-12]
Maximize
Subject to ; ; ; ; .
a)
Subject to ;
b)
Subject to ;
c)
Subject to ;
d)
Subject to ;
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 22
Operation Research Linear Programming Problem
a) Maximize W
Subject to ;
b) Maximize W
Subject to ;
c) Maximize W
Subject to ;
d) Maximize W
Subject to ;
a) Minimize
Subject to ;
b) Minimize
Subject to ;
c) Minimize
Subject to ;
d) Maximize
Subject to ;
a) Minimize S.C ; ;
b) Maximize S.C ; ;
c) Maximize S.C ; ;
This method is similar to the regular simplex method except the selection of variables
to enter and leave the basis. In regular simplex method, first we will select the key column
and then key row. But in revised simplex method, key row is selected first, and then key
column next. In this method the solution starts with optimum but infeasible and works
towards optimality where as regular simplex method starts with a basic feasible solution but
non optimal and works towards optimality.
Step3: Convert the inequality constraints in to equalities by adding slack variables. Obtain
the initial basic feasible solution and put this information in the simplex table.
i) Leaving variable: Find the basic leaving variable corresponding to the most
negative value of Let be the leaving variable.
ii) Entering variable: To find the entering variable, compute the key row (ie)
compute Max . The entering variable is the one having the
maximum ratio consider the rations with negative denomination alone. If
there is no negative denominator then the problem has no feasible solution.
Step6: Follow the regular simplex method to carry out the row operations. Repeat the
procedure until either an optimal solution is obtained.
PG – TRB – Questions
__________________________________________________________________________________
Prepared by D.THIRUMARAN, M.Sc., B.Ed., - 8015106967, C.Mutlur 24