Scribd
Upload
205 views8 pages

Relation Between Linear and Nonlinear Linear Programming Chapter

This chapter discusses the relationship between linear and nonlinear programming problems. It presents a method to convert a linear programming problem into a nonlinear programming problem by raising the variables to a power greater than 1. This conversion allows problems that cannot be solved using the simplex method for linear programming to be solved using nonlinear programming software. Numerical examples demonstrate how the method finds alternate solutions to linear problems that utilize more of the variables.

Uploaded by

ap021
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
205 views8 pages

Relation Between Linear and Nonlinear Linear Programming Chapter

This chapter discusses the relationship between linear and nonlinear programming problems. It presents a method to convert a linear programming problem into a nonlinear programming problem by raising the variables to a power greater than 1. This conversion allows problems that cannot be solved using the simplex method for linear programming to be solved using nonlinear programming software. Numerical examples demonstrate how the method finds alternate solutions to linear problems that utilize more of the variables.

Uploaded by

ap021
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Chapter five

Relation Between lineaR and


NoNliNear ProgrammiNg
5.1 Introduction

L
inear Programming is a technique of expressing real life problems
into mathematical models and solving the models. It is one of the
most powerful and widespread business optimization tools. An
optimization technique capable of solving an amazingly large variety of
business problems, a business objective (e.g. minimize costs of a distribution
system), business restrictions (storage capabilities, transportation volume
restrictions), and costs/revenue (storage costs, transportation costs) are
formulated into a mathematical model.

The controlling variables in mathematical models are the real factors


which we have to be tackled to get the best solution. These factors may be
machine hours, labour components and number of items etc. Quite often, we
face problems in which maximum number of controlling variables are equal
to zero in optimal solution. In real life situation, zero value variables means
no use of that particular variable and it loses its presence. For linear
programming problems having n variables and m constraints, Simplex
Method provides the solution of m basic variables and rest n-m non-
basic variables are considered as zero.

There are relationships between mathematical programming and


regression analysis, experimental design and testing of hypothesis [7,43].
But here, we present the relationship which is totally internal, i.e., the
relationship within mathematical programming. This relationship that helps
to find the better alternate solution of mathematical programming problems,
the method is supported by many examples, some examples are also
discussed in chapter 4 [1]. It also provides n variable solution of
mathematical programming problems having m constraints and n
variables (m < n). where Simplex Method fails to provide such n variable
solution having m constraints only.

5.2 METHOD

Let us consider the Linear Programming Problem

LPPCV: Linear Programming

Max. Z = f(X)

Subject to

aijxj bi (1)

Xj 0 (2)

i=1,2, . . . ,m , j= 1,2, . . . ,n

As the problem (LPPCV ) can easily be solved by Simplex Method


and m variable solution can be obtained.

We convert this linear programming problem into non-linear


programming problem by following way:

NLPPCN: Non-Linear Programming

Max. Z = f(X)

Subject to

aijxjp bi (1)

Xj 0 (2)

i =1,2, . . . ,m, j = 1,2, . . . ,n, p > 1


the problem (NLPPCN) is nonlinear programming problem for different
values of p (p > 1). The solution of non-linear programming problems can
be found by any known methods/softwares, [18,78], Using the solution of
(NLPPCN), we can find the solution of problem (LPPCN).

Following are the few examples that support the method:

5.3 Numerical illustrations

We illustrate the concept with the help of the following examples

Example 5.3.1

Max Z = X 1 + X 2 + X 4

Subject to

X1 + X2 + X3 + X4 4

X1 + 2X2 + X 3 + X5 4

X1 + 2 X2 + X 3 4

Xj 0

By using Simplex Method, the optimal solution is:

X1 = 4, X2 = 0, X3 = 0, X4 = 0 and X5 = 0, giving maximum value


of Z as 4.

The problem has five (5) controlling variables and the Simplex Method
provides the solution (4,0,0,0,0) it means that only one controlling variable
plays active part while others not contributing anything. Now, we change
the linear programming problem into nonlinear programming by taking any
power of p > 1 (say p=2).
Example 5.3.1.1 Non-Linear Programming
Max Z = X 1 + X 2 + X4

Subject to

X12 + X22 + X32 + X42 4

X12 + 2X22+ X32 + X52 4

X12 + 2 X22 + X32 4

Xj 0

The solution of this problem by using LINGO software is obtained as

X1 = 1.154700, X 2 = 1.154700, X 3 = 0, X4 = 1.154700 and X 5 = 0 with


maximum value of Z as 3.461

Now using this solution, we can find the solution to original linear
programming problem given in Example 5.3.1 which is the squares of the
solutions of the Example 5.3.1.1.

X1 = 1.333, X2 = 1.333, X3 = 0, X4 = 1.333, X5 = 0 and value of


Z is 3.999
This is better alternate solution of problem given in Example 5.3.1
consisting of atleast three (3) variables solution instead of one variable
solution as found by Simplex Method.

We can use some higher powers of decision variables to determine


the alternate solutions of mathematical programming problems.
Example 5.3.2

Max Z = X 1 + X 2 + X3 + X4

Subject to

X1 + X2 + X3 + X4 15

7 X1 + 5 X2 + 3 X3 + 2X4 120

3 X1 + 5 X2 + 10 X 3 + 15X4 100

Xj 0

By using Simplex Method optimal solution is:

X1 = 15, X2 = 0, X3 = 0, X4 = 0 with maximum Z= 15

Now, we change the linear programming into nonlinear programming taking


p=3

Example 5.3.2.1

Max Z = X 1 + X 2 + X3 + X4

Subject to

X13 + X 23 + X33+ X43 15

7 X13 + 5 X23 + 3 X33 + 2X43 120

3 X13 + 5 X23 + 10 X33 + 15X43 100

Xj 0

The optimal solution of problem in Example 5.3.2.1 by using LINGO


software is:
X1 = 1.762508, X 2 = 1.645959, X 3 = 1.432785, X4 = 1.285512 and
maximum Z = 6.1267
Proceeding in the similar way, we can find the solution of original
linear programming problem in Example 5.3.2 which is the cube of the
solution of Example 5.3.2.1

X1 = 5.5, X2 = 4.5, X3 = 2.9, X 4 = 2.1 and maximum value of Z as 15

Example 5.3.3

Max Z = X 1 + 2X2 + 3X3 - X4

Subject to

X1 + 2X2 + 3X 3 15

2X1 + X2 + 5X3 20

X1 + 2 X2 + X 3 + X4 10

Xj 0

Optimal solution to the above problem is

X1 = 2.14, X2 =0, X3 = 3.57, X4 = 0 with maximum value of


Z as 15

We change linear programming problem Example 5.3.3 into nonlinear


programming problem Example 5.3.3.1.
Example 5.3.3.1
Max Z = X1 + 2X2 + 3X3 - X 4

Subject to

X14 + 2X24 + 3X 34 15

2X14 + X24 + 5X34 20

X14 + 2 X24 + X34 + X44 10

Xj 0

The optimal solution of this problem by using LINGO software is:

X1 = 1.257433, X2 = 1.257433, X 3 = 1.257433, X4=0 and maximum


value of Z as 7.54

Now by similar way, we can find the solution of original linear


programming problem Example 5.3.3 which is the forth power of the
solution of Example 5.3.3.1.

X1 = 2.5, X2 = 2.5, X3 = 2.5, X 4 = 0 with maximum value Z as 15


This is a better alternate solution of problem given in Example 5.3.3
consisting of three (3) variables solution instead of two.

You might also like