You are on page 1of 28

CHAPTER 6

Integer, Goal, and


Nonlinear Programming
Models

PowerPoint presentation to accompany


Balakrishnan/Render/Stair
based on
Managerial Decision Modeling with Spreadsheets, 3/e
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-1
LEARNING OBJECTIVES
1. Formulate integer programming (IP) models.
2. Set up and solve IP models using Excel’s Solver.
3. Understand the difference between general
integer and binary integer variables.
4. Understand the use of binary integer variables in
formulating problems involving fixed costs.
5. Formulate goal programming (GP) problems and
solve them using Excel’s Solver.

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-2
Introduction

• 3 basic characteristics of LP
• DVs are allowed to have fractional values
• Unique objective function (OF)
• All mathematical expressions
(OF & constraints) had to be linear

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-3
Integer programming
• Extension to LP that solves problems
requiring integer solutions
• Two types of integer variables
• general integer variables
• binary variables.
• General integer variables can take any
non-negative, integer value that satisfies all
constraints in model.
• Binary variables can only take either of two
values: 0 or 1.
based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-4
Integer programming
Pure integer programming problems
All decision variables must have integer
solutions (general integer, binary, or a
combination of the two).

Mixed integer programming problems


Some, but not all, decision variables must
have integer solutions
Non-integer variables can have fractional
optimal values

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-5
Goal programming
• Extension of LP that permits more than
one objective to be stated
• to maximize profit
• to maximize market share
• to maintain full employment
• to minimize costs
• Many of these goals can conflict and
can be difficult to quantify.
based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-6
Nonlinear programming
• Extension to LP in which OF or/and
constraints are nonlinear.
• Price of various products may be a
function of number of units produced.
• As more are made, the selling price per
unit decreases.
max profit = 25X - 0.4X2 + 30Y - 0.5Y2

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-7
Integer Programming

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-8
Solving integer programming

• An IP model has objective function and


constraints identical to LP models.
• No real difference in basic procedure for
formulating an IP model and LP model.
• Only additional requirement in IP model is
one or more of decision variables have to
take on integer values in optimal solution.

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-9
Solving integer programming
• The LP that results from dropping the integer
requirements is called the LP Relaxation of
the IP.
• One-way is solving the LP relaxation problem
and rounding-off the optimal solution, but
• The rounded-off may not be optimal
• The rounded-off may not even be in the
feasible region.
• Unlike LP, optimal solution may not be a
corner point, but any point inside the feasible
region in an IP.
based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-10
An IP example

Write the IP formulation for the problem.

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-11
Harrison Electric
L = number of lamps to make
F = number of ceiling to make

Objective function
Maximize profit = $600L + $700F
subject to

2L + 3F ≤ 12 (wiring hours)
6L + 5F ≤ 30 (assembly hours)
L, F ≥ 0 and integer without this
restriction,
based on it is LP
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-12
LP relaxation
Solving the problem as an LP ignoring the
integer constraints, the optimal solution to
the linear program gives fractional values for
both L and F. The optimal solution to the
linear program is:
L = 3.75 F = 1.5 OFV = 3,300

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-13
Graphical Solution
F Figure 6.1
6–

+ = Integer Valued Point


5–
6L + 5F ≤ 30

4–
Rounded-off IP Solution
(L = 4, F = 2, Infeasible)
3– +

2– + + + + Optimal LP Solution
(L = 3.75, F = 1.50, Profit = $3,300)
1– + + + + 2L + 3F ≤ 12
Nearest Feasible Rounded-off IP Solution
(L = 4, F = 1, Profit = $3,100)
– | | | | | |
0 1 2 3 4 5 6 L
based on
feasible IP solution but is it optimal?
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-14
Integer Solutions
LAMPS (L) CEILING FANS (F) PROFIT ($600L + $700F)
0 0 $ 0
1 0 $ 600
2 0 $1,200
3 0 $1,800
4 0 $2,400
5 0 $3,000
0 1 $ 700
1 1 $1,300
2 1 $1,900
3 1 $2,500
4 1 $3,100  Nearest feasible rounded-off solution
0 2 $1,400
1 2 $2,000
2 2 $2,600
3 2 $3,200  Optimal IP solution
0 3 $2,100
1 3 $2,700
0 4 $2,800 Table 6.1
based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-15
Graphical Solution
F Figure 6.1
6–

+ = Integer Valued Point


5–
6L + 5F ≤ 30

4–
Rounded-off IP Solution
(L = 4, F = 2, Infeasible)
3– + Optimal IP solution
L=3, F=2, Profit=$3,200
2– + + + + Optimal LP Solution
(L = 3.75, F = 1.50, Profit = $3,300)
1– + + + + 2L + 3F ≤ 12
Nearest Feasible Rounded-off IP Solution
(L = 4, F = 1, Profit = $3,100)
– | | | | | |
0 1 2 3 4 5 6 L
based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-16
Solving integer programming

• An important concept
• Optimal IP solution is always a feasible
solution of the same LP problem (i.e., LP-
Relaxation), but not vice versa.
• Optimal IP solution can never be better than
the solution of the same LP problem

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-17
Solving integer programming

• Solution techniques
• Complete enumeration of all feasible points.
• Techniques like Branch & Bound
• Excel-Solver (uses Branch & Bound)

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-18
Solving the Problem

Screenshot 6-1
based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-19
Solving the Problem

Screenshot 6-1
based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-20
Solver Options

From the options


menu, unclick
ignore integer
constraints

Screenshot 6-2A

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-21
Solver Options

Screenshot 6-2B

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-22
Binary Variables
• Only two possible values: 0 or 1
• Xi = 1 if stock i purchased
0 if stock i is not purchased
• We used binary variables in network flow
models (assignment and shortest path)
without specifying as an IP problem.
• The unimodularity property of network models
automatically ensured that optimal values are 0
or 1.

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-23
Sussex County
Sussex County needs to build health care clinics to serve seven communities
(named A to G) in the region. Each clinic can serve communities within a
maximum radius of 30 minutes’ driving time, and a community may be served
more than one clinic. Table 6.3 shows the times it takes to travel between the
seven communities. What is the minimum number of clinics that would be
needed, and which communities should they be located?

TO
FROM A B C D E F G
A 0 15 20 35 35 45 40
B 15 0 35 20 35 40 40
C 20 35 0 15 50 45 30
D 35 20 15 0 35 20 20
E 35 35 50 35 0 15 40
F 45 40 45 20 15 0 35
G 40 40 30 20 40 35 0
Table 6.3

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-24
Sussex County

COMMUNITY COMMUNITIES WITHIN 30 MINUTES


A A, B, C
B A, B, D
C A, C, D, G
D B, C, D, F, G
E E, F
F D, E, F
G C, D, G

Table 6.4

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-25
Sussex County
• Decision Variables
A = 1 if a clinic is located in community A
= 0 if a clinic is not located in community A

Similarly
B (community B),
C (community C),
D (community D),
E (community E),
F (community F),
G (community G)
based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-26
Sussex County
Objective function
Minimize total
number of clinics = A + B + C + D + E + F + G
subject to
A+B+C ≥1 (community A is covered)
A+B+D ≥1 (community B is covered)
A+C+D+G ≥1 (community C is covered)
B+C+D+F+G ≥1 (community D is covered)
E+F ≥1 (community E is covered)
D+E+F ≥1 (community F is covered)
C+D+G ≥1 (community G is covered)
All variables are binary.
based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-27
Solving the Problem

Screenshot 6-4

based on
© 2013 Pearson Education, Inc. publishing as Prentice Hall 6-28

You might also like