You are on page 1of 59

Linear Programming:

Computer Solution
Linear Programming:
Computer Solution
We will show how linear programming
problems can be solved using several
personal computer software packages.

We will also describe how to use a


computer solution result to experiment with
a linear programming model to see what
effect parameter changes have on the optimal
solution, referred to as sensitivity analysis.
Computer Solution
When linear programming was first developed in
the 1940s, virtually the only way to solve a problem
was by using a lengthy manual mathematical solution
procedure called the simplex method.

However, during the next six decades, as computer


technology evolved, the computer was used more and
more to solve linear programming models. The
mathematical steps of the simplex method were
simply programmed in prewritten software packages
designed for the solution of linear programming
problems.
Excel Spreadsheets

Excel can be used to solve linear


programming problems, although the data
input requirements can be more time
consuming and tedious than with a software
package like QM for Windows that is
specifically designed for the purpose.
First open the Microsoft Excel
Click the “Data”
Click the “Solver”
Exhibit 3.1 shows an Excel
spreadsheet set up for our Beaver
Creek Pottery Company example
introduced in “Setting Up and
Editing a Spreadsheet”.
Exhibit 3.1
The values for bowls and mugs and for profit
are contained in cells B10, B11, and B12,
respectively. These cells are currently empty or
zero because the model has not yet been solved.

The objective function for profit,


=C4*B10D4*B11, is embedded in cell B12 shown
in bar. This formula is essentially the same as Z =
40x₁ + 50x₂ , where B10 and B11 represent x₁ and
x₂ and B12 equals Z.
Exhibit 3.1 cont.
The objective function coefficients, 40
and 50, are in cells C4 and D4. Similar
formulas for the constraints for labor and
clay are embedded in cells E6 and E7. For
example, in cell E6 we input the formula
=C6*B10D6*B11.
The <= signs in cells F6 and F7 are for
cosmetic purposes only; they have no real
effect.
Exhibit 3.1 cont.
To solve this problem, first click on the “Data”
tab on the toolbar at the top of the screen and
then click on “Solver” on the far right side of
the Data toolbar. The window Solver Parameters
will appear, as shown in Exhibit 3.2.
Initially all the windows on this screen are
blank, and we must input the objective function
cell, the cells representing the decision
variables, and the cells that make up the model
constraints.
QM for Windows

This is a necessary condition to input


problems into some computer programs,
and specifically QM for Windows, for
linear programming solution.
QM for Windows cont.
We will demonstrate how to use QM for Windows
by solving our Beaver Creek Pottery Company
example. The linear programming module in QM
for Windows is accessed by clicking on “Module”
at the top of the initial window.
This will bring down a menu with all the program
modules available in QM for Windows. By clicking
on “Linear Programming,” a window for this
program will come up on the screen, and by
clicking on “File” and then “New,” a screen for
inputting problem dimensions will appear.
Open the POM-QM
Click “Okay”
Click
“Module”
Click “Linear
Programming”
Click
“File”
Click “New”
Type “Beaver Creek Pottery Company”
Click “Okay”
Click “Solve”
Click “Graph”
Click “Insert”
Click “Scatter”
Click “Scatter with Straight
Lines and Markers”
Click “Right Click”
Click “Select Data”
Click “Edit”
Click “Okay”
Click “Add”
45
40
35
30
25
Labor
20 Clay
15
10
5
0
0 5 10 15 20 25 30 35 40 45
“Ctrl + Shift + Enter
Maximum Z = $40x₁ + $50 x₂

You might also like