Linear Programming

from the companion CD - Chapter 2 of the book:

Balakrishnan, Render, and Stair,
Managerial Decision Modeling with Spreadsheets,
2nd ed., Prentice-Hall, 2007

http://www.stmartin.edu/

Rev. 2.15 by M. Miccio on January 28, 2014

2007 Pearson Education
 Introduction to Linear Programming
A Linear Programming (LP) model seeks
to maximize or minimize a linear function,
subject to a set of linear constraints.
The linear model consists of the following
components:
a set of decision variables.
an objective function.
a set of constraints.

from the companion CD of the book:

Lawrence and Pasternack, Applied Management Science: Modeling, Spreadsheet Analysis, 2
and Communication for Decision Making, 2nd Edition, 2002 John Wiley & Sons Inc.
 Introduction to Linear Programming
The Importance of Linear Programming
Many real world problems lend themselves to linear
programming modeling.
Many real world problems can be approximated by linear
models.
There are well-known successful applications in:
Manufacturing
Marketing
Finance (investment)
Advertising
Agriculture
from the companion CD of the book:
Lawrence and Pasternack, Applied Management Science: Modeling, Spreadsheet Analysis, 3
and Communication for Decision Making, 2nd Edition, 2002 John Wiley & Sons Inc.
 Introduction to Linear Programming
The Importance of Linear Programming
There are efficient solution techniques that solve linear
programming models.
The output generated from linear programming
packages provides useful what if analysis.

from the companion CD of the book:

Lawrence and Pasternack, Applied Management Science: Modeling, Spreadsheet Analysis, 4
and Communication for Decision Making, 2nd Edition, 2002 John Wiley & Sons Inc.
 Introduction to Linear Programming
Assumptions of the linear programming model
The parameter values are known with certainty.
The objective function is linear.
z = c1 x1 + c2 x2 +.....+ cj xj +.....+ cn xn = cT x
where cT = ( c1, c2,...ci,..., cn ) is the cost coefficient vector
The constraints are linear.
ai1 x1 + ai2 x2 +.....+ aij .xj+.....+ ain xn = bi
ak1 x1 + ak2 x2 +.....+ akj .xj+.....+ akn xn bk
ar1 x1 + ar2 x2 +.....+ arj .xj+.....+ arn xn br
There are no interactions between the decision variables (the
additivity assumption).
The continuity assumption: Variables can take on any value within
a given feasible range.
from the companion CD of the book:
Lawrence and Pasternack, Applied Management Science: Modeling, Spreadsheet Analysis, 5
and Communication for Decision Making, 2nd Edition, 2002 John Wiley & Sons Inc.
 LP Terminology
NON-NEGATIVITY CONSTRAINT
The decision variables are non-negative, i.e., they are
NOT allowed to take on negative values

For instance, for a 2-D LP model:

X2

The non-negativity
constraint

X1

from the companion CD of the book:

Lawrence and Pasternack, Applied Management Science: Modeling, Spreadsheet Analysis, 6
and Communication for Decision Making, 2nd Edition, 2002 John Wiley & Sons Inc.
 LP Terminology
FEASIBLE REGION
The set in the space of the decision variables of all points
that satisfy all the constraints of the model at the same
time.

7
 LP Terminology
Feasible points in the feasible region:
three types are possible
X2
1000 The Plastic constraint
2X1+X2 1000
700 Total production constraint:
X1+X2 700 (redundant)
500
Infeasible
Production mix
constraint:
Production Feasible X1-X2 350
Time
3X1+4X22400
X1
500 700 8
Interior points Boundary points Extreme points
 LINEAR PROGRAMMING
Application to the Food Industry
(Example of a Mixing Problem) - I

A brewery has received an order for 100 gal of beer with the
special constraint that the beer must contain 4% alcohol by
volume and it must be supplied immediately.
The brewery wishes to fill the order, but no 4% beer is now
in stock. It is decided to mix two beers now in stock to give
the desired final product. One of the beers in stock (Beer A)
contains 4.5% alcohol by volume and is valued at $0.32 per
gallon. The other beer in stock (Beer B) contains 3.7%
alcohol by volume and is valued at $0.25 per gallon. Water
(W) can be added to the blend at no cost.
What volume combination of the two beers in stock with
water, including at least 10 gal of Beer A, will give the
minimum ingredient cost for the 100 gal of 4% beer?

9
 LINEAR PROGRAMMING
Application to the Food Industry
(Example of a Mixing Problem) - II

Three independent variables (decision variables)

are considered, i.e., amount of Beer A (VA), amount
of Beer B (VB), and amount of water (Vw).

This example is greatly simplified because only a

few simple constraints are involved.

When a large number of possible choices is

involved, the optimum set of choices may be far from
obvious, and a solution by Linear Programming may
be the best way to approach the problem.

10
 LINEAR PROGRAMMING
Application to the Food Industry
(Example of a Mixing Problem) - III
Mathematical Model
Objective function
z = 0.32 VA + 0.25 VB min!

Problem Constraints
1.VA + VB + Vw = 100
2.VA 10
3.(1 - 0.045) VA + (1 - 0.037) VB + Vw = (1 - 0.04)100
[3 0.045VA + 0.037VB = 4]

Non-negativity Restriction Brewery (after substitution)

1.VA 0 VB 0 Vw 0
Modified Constraints
By taking Vw = 100 (VA + VB) from eq.1 and
Optimal Solution
VA = 37.5 gal; replacing:
1.VA 10
VB = 62.5 gal;
2.0.955 VA + 0.963 VB + 100 (VA + VB) = 96
z = 27.63 $/(100 gal)
2. 0.045 VA + 0.037 VB = 4
11
 Steps in Developing
a Linear Programming Model
1) Formulation
a) Understanding the problem
b) Identification of decision variables
c) Representation of the objective function
d) Representation of constraints

2) Solution

3) Interpretation and Sensitivity Analysis 12

 Special Situation in LP

1. Redundant Constraints: do not affect

the feasible region

Example: x1 < 10
x1 < 12
The second constraint is redundant
because it is less restrictive.

13
 Special Situation in LP

2. Infeasibility: when no feasible solution

exists (there is no feasible region)

Example: x1 < 10
x1 > 15

14
 Infeasible Model
No point, simultaneously,
lies both above line and
below lines and .

3 1

from the companion CD of the book:

Lawrence and Pasternack, Applied Management Science: Modeling, Spreadsheet Analysis, 15
and Communication for Decision Making, 2nd Edition, 2002 John Wiley & Sons Inc.
 Special Situation in LP
3. Alternate Optimal Solutions: when
there is more than one optimal solution

Max 2T + 2C C
Subject to: 10

2T
T + C < 10

+
All points on

2C
T < 5

=
Red segment

20
6
C< 6 are optimal
T, C > 0
NB:
The objective function eq. is
parallel to the eq. of one 0
0 5 10 T 16
constraint
 Special Situation in LP

4. Unboundness: when nothing prevents

the solution from becoming infinitely
large
C

lut n
so tio
ion
Max 2T + 2C

of rec
Subject to:

Di
2
2T + 3C > 6
T, C > 0 1

0
0 1 2 3 T 17
 Special Situation in LP

4. Unboundness (2nd example in 2D):

when nothing prevents the solution from
becoming infinitely large

of
t i on
c n
i re lutio
D o
s
18
 Special Situation in LP
5. Decision variable unrestricted in sign:

x j
x j x 'j x "j
with x 'j 0; x "j 0

19
 Sensitivity Analysis
of the Optimal Solution

Is the optimal solution sensitive to changes in

input parameters?

Possible reasons for asking this question:

Parameter values used were only best estimates.
Dynamic environment may cause changes.
What-if analysis may provide economical and
operational information.

from the companion CD of the book:

Lawrence and Pasternack, Applied Management Science: Modeling, Spreadsheet Analysis, 20
and Communication for Decision Making, 2nd Edition, 2002 John Wiley & Sons Inc.
 Example:
Sensitivity Analysis of Objective Function Coefficients
1000 X2

M
Ma ax
x 3 4X
.75 1 +
X 5X M
1 +
5X 2 ax
2 8X
1
+
600 5X
2

X1

500 800
from the companion CD of the book:
Lawrence and Pasternack, Applied Management Science: Modeling, Spreadsheet Analysis, 21
and Communication for Decision Making, 2nd Edition, 2002 John Wiley & Sons Inc.