Scribd
Upload
95 views

Solution of Linear Programming Problems With Matlab

This document introduces linear programming problems and how to solve them using MATLAB's linprog command. It provides notation for linear constraints and defines linear programming problems as optimization problems with linear objective functions and constraints. An example winemaker problem is presented and solved using linprog. Finally, three practice problems are presented and the reader is instructed to set them up in the proper form to solve with linprog.

Uploaded by

Mahfuzulhoq Chowdhury
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
95 views

Solution of Linear Programming Problems With Matlab

This document introduces linear programming problems and how to solve them using MATLAB's linprog command. It provides notation for linear constraints and defines linear programming problems as optimization problems with linear objective functions and constraints. An example winemaker problem is presented and solved using linprog. Finally, three practice problems are presented and the reader is instructed to set them up in the proper form to solve with linprog.

Uploaded by

Mahfuzulhoq Chowdhury
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/ 3

Solution of Linear Programming Problems with Matlab

Notation

• The transposition operation is denoted by a superscript T (apostrophe in Matlab),


   T  
1 1 " #T 1 4
T     1 2 3  
[1, 2, 3] =  2  ,  2  = [1, 2, 3], =  2 5 .
4 5 6
3 3 3 6

• Given two (row or column) vectors a and b with components a1 , . . . , an and b1 , . . . , bn ,


the notation
a ≤ b or a ≥ b
is a shorthand notation for

ai ≤ bi or ai ≥ bi for all 1 ≤ i ≤ n.

Definition 1. Let f be a column vector of length n, b a column vector of length m, and


let A be a m × n–matrix.
A linear program associated with f , A, and b is the minimum problem

min f T x (1)

or the maximum problem


max f T x (2)
subject to the constraint
Ax ≤ b. (3)

Note that x is a column vector of length n.


The general version of a linear program may involve inequality constraints as well as
equality constraints:

Definition 2. Let f be a column vector of length n, b a column vector of length m,


beq a column vector of length k, and let A and Aeq be m × n and k × n matrices, respec-
tively.
A linear program associated with f , A, b, Aeq , beq is the minimum problem (1) or the
maximum problem (2), subject to the inequality constraint (3) and the equality constraint

Aeq x = beq . (4)

Example. The winemaker example led us to the following problem:

12x1 + 7x2 = max,

1
subject to

2x1 + x2 ≤ 10, 000


3x1 + 2x2 ≤ 16, 000
x1 ≥ 0,
x2 ≥ 0.

If we define    
" # 10, 000 2 1
12  16, 000   3 2 
   
f= , b= , A= ,
7  0   −1 0 
0 0 −1
this problem can be identified with the linear programming maximum problem associated
with f , A, b. Likewise it can be identified with the linear programming minimum problem
associated with −f , A, b.

Solution of linear programming minimum problems with Matlab


Matlab provides the command linprog to find the minimizer (solution point) x of a linear
programming minimum problem. Without equality constraint the syntax is

x=linprog(f,A,b)

If you also want to retrieve the minimal value fmin = minx (f T x), type

[x,fmin]=linprog(f,A,b)

If inequality and equality constraint are given, use the commands

x=linprog(f,A,b,Aeq,beq)

or

[x,fmin]=linprog(f,A,b,Aeq,beq)

Let’s solve our winemaker problem:

>> f=[-12;-7];b=[10000;16000;0;0];A=[2 1;3 2;-1 0;0 -1];


>> [x,fopt]=linprog(f,A,b)
Optimization terminated successfully.

x =

1.0e+003 *

3.99999999989665

2
2.00000000013951

fopt =

-6.199999999973631e+004
This is the answer found in the class notes. The solution point is (4000, 2000), and the
maximum profit is $6, 2000.

Practice Problems

In each of the following problems first identify vectors and matrices such that the op-
timization problem can be written in the form of Definitions 1 or 2. Then use the linprog
command to solve the linear program.

Problem 1.
x1 + x2 = max
subject to
2x1 + x2 ≤ 29,
x1 + 2x2 ≤ 25,
x1 ≥ 2,
x2 ≥ 5.

Problem 2.
x1 + x2 + x3 + x4 + x5 = max
subject to
x1 + x2 ≤ 100,
x3 + x4 ≤ 70,
x2 + x3 + 2x4 + 5x5 ≤ 250,
xj ≥ 0 (1 ≤ j ≤ 5).

Problem 3.
x1 + x2 + x3 + x4 + x5 = min
subject to
x1 + x2 = 100,
x3 + x4 = 70,
x2 + x3 + 2x4 + 5x5 = 250,
xj ≥ 0 (1 ≤ j ≤ 5).

You might also like