Introduction to Optimization

(Ch4-1)

March 22, 2022

Jin Young Choi

Department of Industrial Engineering
Ajou University

IE-AJU Jin Young Choi

 Today’s Outline

 Brief Review of Lecture Ch3-4

 The essence of the simplex method

 Setting up the simplex method

IE-AJU Jin Young Choi

 Review of Lecture Ch3-4
 One more Example of LP formulation
– Distributing Goods through a Dist. Net

 Formulating and solving LP models

– Solution procedure
• Using the Excel solver
• Using Lingo

IE-AJU Jin Young Choi

 The essence of the simplex method
 Why Geometric ?
– The simplex method is an algebraic procedure. However, its
underlying concepts are geometric
– Understanding these geometric concepts provides a strong
intuitive feeling for how the simplex method operates and what
makes it so efficient

 Corner-point solutions
– A constraint boundary is a line that forms the boundary permitted
by the corresponding constraint
– The points of intersections of constraint boundaries are the
corner-point solutions of the problem
• Corner-point feasible (CPF) solutions
• Corner-point infeasible solutions

IE-AJU Jin Young Choi

  Properties of CP solutions and CPF solutions
For a linear programming problem with n decision variables
1. Each of its corner-point solutions lies at the intersection of (at
least) n constraint boundaries
2. Two CPF solutions are adjacent to each other if they share n-1
constraint boundaries
3. The two adjacent CPF solutions are connected by a line
segment that lies on these same shared constraint boundaries
• Such a line segment is referred to as an edge of the feasible
region

IE-AJU Jin Young Choi

  Optimality test
Consider any linear programming problem that possesses at
least one optimal solution
If a CPF solution has no adjacent CPF solutions that are better,
then it must be an optimal solution

<a sketch of proof> If it is not optimal, there exists an adjacent CPF

solution that are better  contradiction

 One of the key solution concepts

– The simplex method only checks CPF solutions to find an
optimal solution

IE-AJU Jin Young Choi

  Applying the simplex method to the Wyndor prob.
(from a geometric viewpoint)

– Initialization (of the initial CPF solution)

• Choose (0, 0) as the initial CPF solution to examine
• This is a convenient choice because no calculations are required to
identify this CPF solution

– Optimality test
• Adjacent CPFs (0, 6) and (4, 0) have z = 30 and 12, respectively
• (0, 0) is not optimal

IE-AJU Jin Young Choi

 – Iteration 1
• Consider the two edges that emanate from (0, 0)
• Choose to move along the edge that leads up the x2 axis
– Z=3x1+5x2  moving up the x2 axis increases Z at a faster rate
• Stop at the first new constraint boundary 2x2 = 12
• Solve x1 = 0 and 2x2 = 12  (0, 6)

– Optimality test
• Adjacent CPFs (0, 0) and (2, 6) have z = 0 and 36, respectively
• (0, 6) is not optimal

IE-AJU Jin Young Choi

 – Iteration 2
• Consider the two edges emanating from (0, 6)
• Choose to move along the edge that leads to the right
– Moving along this edge increases Z
• Stop at the first new constraint boundary 3x1+2x2=12
• Solve for the intersection of 3x1+2x2=12 and 2x2=12  (2, 6)

– Optimality test
• Z = 36 at (2, 6)
• Adjacent CPFs (0, 6) and (4, 3) have z = 30 and 27, respectively
• No adjacent CPFs are better
• (2, 6) is an optimal solution

IE-AJU Jin Young Choi

  The key solution concepts
1. (The relationship between optimal solutions and CPF solutions)
The simplex method focuses solely on CPF solutions. For any problem
with at least one optimal solution, finding one requires only finding a
best CPF solution

2. (The flow of the simplex method)

The simplex method is an iterative algorithm with the following
structure
Initialization

Optimality Test

no
Optimal Iteration
yes

Stop
IE-AJU Jin Young Choi
 3. (How to get started)
The simplex method chooses the origin (all decision variables equal to
zero) to be the initial CPF solution

* Choosing the origin commonly is possible when all the decision

variables have nonnegativity constraints

4. (The choice of a better CPF solution at each iteration)

The simplex method always chooses a CPF solution that is adjacent to
the current one

* Given a CPF solution, it is much quicker computationally to gather

information about its adjacent CPF solutions than about other CPF
solutions

IE-AJU Jin Young Choi

 5. (Which adjacent CPF solution to choose at each iteration)
The simplex method chooses to move along the edge with the largest
rate of improvement in Z (among the edges with a positive rate of
improvement in Z)

6. (How the optimality test is performed efficiently)

The optimality test consists simply of checking whether any of the
edges give a positive rate of improvement in Z. If none do, then the
current CPF solution is optimal

IE-AJU Jin Young Choi

 Setting up the simplex method
 Translating into an algebraic procedure
– It is necessary to translate the conceptually geometric procedure
into a usable algebraic procedure
– The algebraic procedure is based on solving systems of
equations
– Therefore, the first step in setting up the simplex method is to
convert the functional inequality constraints to equivalent
equality constraints
• By introducing slack variables

IE-AJU Jin Young Choi

  The augmented form
– It is obtained from the original form by introducing some
supplementary variables needed to apply the simplex method

IE-AJU Jin Young Choi

  Terminology for the augmented form
– An augmented solution
• A solution of the augmented problem
• Ex) (3, 2) is a solution of the original prob. Augmenting (3, 2) yields
the augmented solution (3, 2, 1, 8, 5)

– A basic solution
• An augmented corner-point solution
• It might be feasible or not.
• Ex) Augmenting a CP (0, 6)  (0, 6, 4, 0, 6)  feasible
Augmenting a CP (4, 6)  (4, 6, 0, 0, -6)  infeasible

– A basic feasible solution (BFS)

• An augmented CPF solution
• Ex) Augmenting a CPF solution (0, 6)  (0, 6, 4, 0, 6)  feasible

IE-AJU Jin Young Choi

 – Correspondence between basic solution (BF solutions) and
corner-point solutions (CPF solutions)
• Basic solutions (BF solutions) include the value of the slack variables
• For any basic solution, the corresponding corner-point solution is obtained
simply by deleting the slack variables

 Algebraic properties of basic solutions and BF solutions

– For the augmented form of the example
• # of variables – # of equations = 5 – 3 = 2
• 2 degrees of freedom
– Any two variables can be chosen to be set equal to any arbitrary value
in order to solve the three equations
– Nonbasic variables  these two of the variables
– Basic variables  the other three variables
• Basic solution
– the simultaneous solution of the three equations after setting the values
of the two nonbasic variables

IE-AJU Jin Young Choi

  Properties of a basic solution
– Each variable is designated as either a nonbasic variable or a
basic variable
– The # of basic variables = the # of functional constraints
• The # of nonbasic variables = the total # of variables – the # of
functional constraints
– The nonbasic variables are set equal to zero. (in the simplex
method)
– The values of the basic variables are obtained as the
simultaneous solution of the system of equations
• The set of basic variables is referred to as the basis
– If the basic variables satisfy the nonnegativity constraints, the
basic solution is a BF solution

IE-AJU Jin Young Choi

  Adjacency of two BF solutions
– Two BF solutions are adjacent iff all except one of their nonbasic
variables are the same
– Ex)
• CPF solutions (0, 0) and (0, 6) are adjacent
• Their augmented solutions are (0, 0, 4, 12, 18) and (0, 6, 4, 0, 6)
– They are automatically adjacent
• Then, their nonbasic variables (x1, x2) and (x1, x4) are the same with
just one exception x2 (x2 has been replaced by x4)
• Consequently, moving from (0, 0, 4, 12, 18) to (0, 6, 4, 0, 6) involves
switching x2 from nonbasic to basic and vice versa for x4

– Moving from the current BFS to an adjacent one involves

• Switching one variable from nonbasic to basic and vice versa for
one other variable, and then
• Adjusting the values of the basic variables to continue satisfying the
system of equations

IE-AJU Jin Young Choi

  Starting the simplex method
– It is convenient to consider and manipulate the objective function
equation as the new constraint equations as follows
Max Z
s.t. Z  3x1  5 x2  0 (0)
x1  x3  4 (1)
Assuming bi  0 for all i
2 x2  x4  12 (2)
3x1  2 x2  x5  18 (3)
x j  0 for all j

 When using Eqs. (1) – (3) to obtain a basic solution, we use Eq.
(0) to solve for Z at the same time

IE-AJU Jin Young Choi

 Announcements
 On March 24, 2022
– The algebra of the simplex method
– Reading assignment: Sec.4.3

IE-AJU Jin Young Choi