LINEAR PROGRAMMING:

AN ALGEBRAIC APPROACH
Week 8
The Simplex Method
The method of corners is not suitable for solving linear programming
problems when the number of variables or constraints is large. Its
major shortcoming is that a knowledge of all the corner points of the
feasible set S associated with the problem is required.
Thus we need to reduce the number of points to be inspected. One
technique is the simplex method, which was developed in the late
1940s by George Dantzig and is based on the Gauss–Jordan elimination
method.
The simplex method is readily adaptable to the computer, which makes
it suitable for solving linear programming problems involving large
numbers of variables and constraints.
An Iterative Method
The simplex method is an iterative procedure: it is repeated
over and over again.
Beginning at some initial feasible solution (a corner point of
the feasible set S, usually the origin), each iteration brings us
to another corner point of S, usually with an improved (but
certainly no worse) value of the objective function.
The iteration is terminated when the optimal solution is
reached (if it exists).
4.1 The Simplex Method:
Standard Maximization Problems
A Standard Maximization Problem
A standard maximization problem is one in which
1. The objective function is to be maximized.
2. All the variables involved in the problem are nonnegative.
3. All other linear constraints may be written so that the expression involving the variables is less
than or equal to a nonnegative constant.

Example.

The optimal solution to the problem occurs at the corner point C(3, 2).
Using the Simplex Method
First replace the system of inequality constraints with a system of equality constraints.
2x + 3y  12 is converted into the equation 2x + 3y + u = 12
The variable u is called a slack variable.
2x + y  8 is converted into the equation 2x + y + v = 8
Finally, rewriting the objective function in the form -3x - 2y + P = 0.
We are led to the following system of linear equations:

From among all the solutions of the system for which x, y, u, and v are nonnegative (such solutions
are called feasible solutions), determine the solution(s) that maximizes P.
The Augmented Matrix

Observe that each of the u-, v-, and P-columns of the augmented matrix (4) is a unit
column. The variables associated with unit columns are called basic variables; all other
variables are called non-basic variables.
The configuration of the augmented matrix suggests that we solve for the basic variables u,
v, and P in terms of the non-basic variables x and y, obtaining
Basic Solution
A particular solution is obtained by letting x = 0 and y = 0.
In fact, this solution is given by x = 0, y = 0, u = 12, v = 8, P = 0.
Such a solution, obtained by setting the non-basic variables equal to zero, is
called a basic solution of the system.
This particular solution corresponds to the corner point A(0, 0) of the
feasible set associated with the linear programming problem.
If the value of P cannot be increased, we have found the optimal solution to
the problem at hand.
Continuing our quest for an optimal solution, our next task is to determine
whether it is more profitable to increase the value of x or that of y
(increasing x and y simultaneously is more difficult).
Is it better to hold x = 0 or y = 0?
How Much Can x be Increased
While Holding y = 0?
Setting y = 0,

The first equation and the nonnegativity of u imply that x cannot exceed 12/2 or 6.
The second equation and the nonnegativity of v imply that x cannot exceed 8/2 or 4. Thus,
we conclude that x can be increased by at most 4.
Now, if we set y = 0 and x = 4, we obtain the solution
x = 4, y = 0, u = 4, v = 0, P = 12
which is a basic solution to system, this time with y and v as non-basic variables.
We have to find an augmented matrix that is equivalent to the matrix and has a
configuration in which the x-column is in the unit form
Pivot
Continuing
Consider the pivot column, use row operations to change the pivot element into 1
and non-pivot elements into 0s.
Continue to find the next pivot column, pivot row, and pivot element with similar
procedure as before.

Then, again, change the pivot element into 1 and non-pivot elements into 0s.
We stop the iteration when there are no negative entries in the last row, which
means the solution is optimal and P cannot be increased further.
Simplex Method First Step:
Set Up Initial Simplex Tableu
Example
Set up the initial simplex tableau for the linear programming problem
The Simplex Method
Example
Applied Example
Problems with Multiple Solutions and
Problems with No Solutions
A linear programming problem will have infinitely many solutions if and
only if the last row to the left of the vertical line of the final simplex
tableau has a zero in a column that is not a unit column.
A linear programming problem will have no solution if the simplex
method breaks down at some stage.
For example, if at some stage there are no nonnegative ratios in our
computation, then the linear programming problem has no solution.
4.2 The Simplex Method:
Standard Minimization Problems
Minimization with  Constraints
By changing the problem to a standard maximization problem which satisfies three conditions:
1. The objective function is to be maximized.
2. All the variables involved are nonnegative.
3. Each linear constraint may be written so that the expression involving the variables is less than
or equal to a nonnegative constant.

Example.

Minimizing the objective function C is equivalent to maximizing the objective function P = -C.
Maximize P = 2x + 3y subject to the given constraints.
Standard Minimization Problems
A class of linear programming problems which is characterized by the
following conditions:
1. The objective function is to be minimized.
2. All the variables involved are nonnegative.
3. All other linear constraints may be written so that the expression
involving the variables is greater than or equal to a constant.

Each maximization linear programming problem is associated with a

First, write the tableu for the primal problem:

Next, we interchange the columns and rows of the foregoing tableau and head the two columns of the
resulting array with the two variables u and √, obtaining the tableau:

.
Primal Problem vs Dual Problem
Simplex Method for Minimization
Duality
Example
Assignment
Solve the following problems in a group of at most 2 persons.
1. Read Tan textbook page 228-231: solving linear programming with
Excel. Then solve problems 1 and 4 in page 231 using Excel.
2. Read Tan textbook page 246-248: solving linear programming with
Excel. Then solve problems 6 and 8 in page 248 using Excel.

Submit 4 excel files for each problem to ma1203itb@gmail.com.