Linear Programming Lecture Notes - Penn State Personal Web Server
Linear Programming Lecture Notes - Penn State Personal Web Server
Linear Programming Lecture Notes - Penn State Personal Web Server
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
CHAPTER 6<br />
Simplex Initialization<br />
In the previous chapter, we introduced the Simplex Algorithm and showed how to manipulate<br />
the A, B and N matrices as we execute it. In this chapter, we will discuss the issue<br />
of finding an initial basic feasible solution to start execution of the Simplex Algorithm.<br />
1. Artificial Variables<br />
So far we have investigated linear programming problems that had form:<br />
max c T x<br />
s.t. Ax ≤ b<br />
x ≥ 0<br />
In this case, we use slack variables to convert the problem to:<br />
max c T x<br />
s.t. Ax + Imxs = b<br />
x, xs ≥ 0<br />
where xs are slack variables, one for each constraint. If b ≥ 0, then our initial basic feasible<br />
solution can be x = 0 and xs = b (that is, our initial basis matrix is B = Im). We have<br />
also explored small problems where a graphical technique could be used to identify an initial<br />
extreme point of a polyhedral set and thus an initial basic feasible solution for the problem.<br />
Suppose now we wish to investigate problems in which we do not have a problem structure<br />
that lends itself to easily identifying an initial basic feasible solution. The simplex algorithm<br />
requires an initial BFS to begin execution and so we must develop a method for finding such<br />
a BFS.<br />
For the remainder of this chapter we will assume, unless told otherwise, that we are<br />
interested in solving a linear programming problem provided in Standard Form. That is:<br />
⎧<br />
⎪⎨ max c<br />
(6.1) P<br />
⎪⎩<br />
T x<br />
s.t. Ax = b<br />
x ≥ 0<br />
and that b ≥ 0. Clearly our work in Chapter 3 shows that any linear programming problem<br />
can be put in this form.<br />
Suppose to each constraint Ai·x = bi we associate an artificial variable xai . We can<br />
replace constraint i with:<br />
(6.2) Ai·x + xai = bi<br />
91