You are on page 1of 58

Chapter 3

Linear
Linear Programming
Programming
Models
Models
1


Introduction to Linear
Programming

(Linear Programming model)


(a set of linear constraints)
(maximize) (minimize)
(objective function)
:
(A set of decision variables)
(An objective function)
(A set of constraints)
2


Introduction to Linear
Programming




Manufacturing
Marketing
Finance (investment)
Advertising
Agriculture
3


Introduction to Linear
Programming



what if


Assumptions for Linear
Programming
(certainty)

(constant returns to scale)
1 $4
3Hrs

500
$4*500=$2000 3*500
=1,500Hrs

(Continuity)


The Galaxy Industries
Production Problem
Galaxy :
Space Ray.
Zapper.

(Resources)
1000 (special
plastic)
40 (40 hrs of
production time per week)
6


The Galaxy Industries
Production Problem
(Marketing requirement)
700
Space Rays Zappers 350

(Technological inputs) (Table


2.2)
Space Rays 2 pounds
3
7
Zappers 1pound 4


The Galaxy Industries
Production Problem

Space Ray (profit)


$8 Zappers (profit) $5
Space Ray
Zapper
:
8(450) + 5(100)
Space Rays = 450 dozen
Zapper
= 100 dozen
Profit
= $4100 per week
8

Management is
seeking a production
schedule that will
increase the

A linear programming model


can provide an insight and an

10


The Galaxy Linear
Programming Model
(Decisions variables):
X1 = Space Rays
X2 = Zappers
(Objective Function):

11


The Galaxy Linear
Programming Model
Max 8X1 + 5X2

( )

subject to
2X1 + 1X2 1000 (
,Plastic)
3X1 + 4X2 2400 (
,Production Time)
X1 + X2 700
( ,Total
production)
12


Graphical Analysis of Linear
Programming

The set of all points that


satisfy all the constraints of
the model
is called a
FEASIBLE
REGION
13

(graphical
presentation)
(all the
constraints)
(objective
function)
(three types of

14


Graphical Analysis the
Feasible Region
X2

The non-negativity constraints


( )
X1

15


Graphical Analysis the
Feasible Region
X2

1000

Plastic
2X1+X2 1000

700
500

Total production
X1+X2 700 ( )

Infeasible

Feasible
Production
Time
3X1+4X2 2400
500

700

X1

16

(p. 67~68)
Graphical Analysis the
Feasible Region
X
2

1000
700
500

Plastic
2X1+X2 1000
Total production
X1+X2 700 ( )

Infeasible
Mix

X1-X2 350

Feasible
Production
Time
3X1+4X2 2400
500

700

X1

Interior points. Boundary Extreme point


points.
(feasible
points)
17


Solving Graphically for an
Optimal Solution

18

(p.71)
The search for an optimal
solution

profit , say profit = $1,250.


increase the profit, if possib
X2

1000

700

continue until it becomes infeasible

500

Optimal
Profit
=$4360

Profit
=$125
0

X1
500

19

(p.69)
Summary of the optimal solution
Space Rays X1 * = 320 dozen
Zappers
Profit

Exceldozen
X2 * = 360

Z * = $4360

(plastic)
(production hours).
2X1 + 1X2 = 1000
3X1 + 4X2 = 2400
Time)

( ,Plastic)
( ,Production

(Binding Constraints):

20

(p.70~71)
Summary of the optimal solution
(Total production) 680 (not 700 )

X1 +
X2 Rays
= 680
< 700
Space

( )
X1 - X2 = -40 < 350
( )

Zappers
40

700-680=20
350-(-40) = 390

(Non-Binding Constraints)

(Slack)
21

(p.72)
Extreme points and optimal
solutions

(
,True/False)

3X1+4X2 =
2400
X1 = 0 (0,600)

2X1+ X2 = 1000
3X1+4X2 = 2400

(320,360)

2X1+ X2 =
(450,100)
1000
X1-X2 = 350

22


Multiple optimal solutions

X2=(0,600)
2

X=X1+(1-)X2
, [0,1]

X1=(350,0) 1

Z
23


The Role of
Sensitivity Analysis
of the Optimal Solution
(p.75)

.. (What-if)
.
24

(1)
Sensitivity Analysis of

Objective Function
Coefficients.

(Range of Optimality) (p.76)



?
(p.77)


.
25


Sensitivity Analysis of
Objective Function
Coefficients.
X2

1000

(320,360
)

5X
2

(0,600)

8X

600

C1 =2

C1
83.75

ax
M

M
Ma ax
x3 4
.7 X1
5X +
1 + 5X
5X2
2
(0,600)

M
(320,360)ax 2X

(320,360)

+5
X2

X1
500

800

26


Sensitivity Analysis of
Objective Function
Coefficients.
X
1000
2

1
8X
ax
M

C1 810
(320,360)

M
10
ax

2
5X

600 M

X1

ax
3.
75
X

+
+ (320,360)
5X
2
5X

C1 :
[3.75, 10]
C2
: [4, 10.67]
(Can
you find it ?)

400

600

800

X1

27

Reduced cost
(p.78)

Xj =0 RCj
(-Zj)
(Xj >0)
C1=2 X*=(0,600) X1=0
C1=3.75 X*=(320,360)
X1=320>0
RC1 =-Z1=-(3.75-2)=-1.75

RCj Xj
(Xj=1)
28


(p.79)
X2

1000
Ma
x

3.
75
X

X1
1
+

5X

(0,600)
Z=300600
0

X1=1 (
X1=0X1=1)
Z=2998.25-3000
= -1.75
RC1 =-1.75

(1,599.2
5)
Z=2998.
25
Max

2X
1 +
5X
2

X1
500

800

29

(2)
(p.78) Sensitivity Analysis of
Right-Hand Side Values


?
( )

30


Sensitivity
Analysis of
Right-Hand Side
Values

(Binding Constraints)

(Non-Binding Constraints)
(slack)
(surplus)
(Shadow Price)

31

Shadow Prices
(p.80)

32

Shadow Price
graphical demonstration

Plastic

X2

2X 1
+

2X 1

1000

01
10
<=
00
10
<=

1x 2

1x 2

+
500

X*=(320,36

0)
X*
(320,360)(320.8,359.4
Z*=
=(320.8,359.4)
$4360 Z* = $4363.4)

Producti
on time
(
10001001)

Shadow price =
4363.40 4360.00
= 3.40
X1
500

33

Range of
Feasibility (p.81)


.

Change in objective
value =
[ Shadow price]*[ Change
in the right hand side value]
34

Range of
Feasibility (p.81)

Plastic

X2

2X 1
+

Binding

10
<=
00

Total
Production

500
X1 + X2
700
Production time

1x 2

1000

Total Production

(New Binding Constraint)

X1
500

35

Plastic

Range of
Feasibility

X2

2X 1
+
0
00
1

Total Production

X1+X2 700

1x 2

1000


= 2X1 + 1X2
=2*(400)+300=110
0
X1+ X2 = 700
3X1+4X2 = 2400

X*=(400,300)

600

Production time

3X1+4X2 2400
500

X1
36

Range of
Feasibility
X2

Infeasible 1000
solution

600

X1 = 0

3X1+ 4X2 =
2400
X1 = 0

X*=(0,600)

Plastic
2X1 + 1X2
1000

=2X1 + 1X2
= 2*(0)+1*600=600

500

Production time

3X1+4X2 2400

X1
37

The correct
interpretation of shadow prices
(p.83)
(Sunk costs):
- Shadow Price

1000 $3 Total Cost =


$3000
Production Time $20/hr Total Cost
=$20*40=$800
Production
Time $3000+$800=$3800

$3800

38

The correct
interpretation of shadow prices
(p.83)
(Included costs):
Shadow Price

$3 =$3.4

$6.8(
)
Production Time $0.33/min (or $20/hr)
Production Time =$0.4
Production Time
$0.73

p.84 2.5
39

(3) (p.84)
Other Post - Optimality
Changes
(Addition of a constraint)

Yes, the solution is still optimal


No, re-solve the problem (the new
objective function is worse than the
original one)

(Deletion of a constraint)

Yes, re-solve the problem (the new


objective function is
better than
the original one)
No, the solution is still optimal

40

(p.84)
Other Post - Optimality
Changes

(Deletion of a variable)
0
Yes, the solution is still optimal
No, re-solve the problem (the new
objective function is worse than the
original one)

(Addition of a variable)
(Net Marginal Profit)
41

(p.85)
Other Post - Optimality
Changes
X3=
3lb
5min
$10

Max 8X1 + 5X2+ 10X3


( )
subject to
2X1 + 1X2 + 3X3 1000 ( ,Plastic
,Shadow Price = $3.4)
3X1 + 4X2 + 5X3 2400 ( ,Production
Time, SP = $0.4)
X1 + X2 +X3 700
( ,Total
production, SP = $0)
X1 - X2
350
( , SP = $0)
Xj> = 0,
j = 1,2,3
( ,Nonnegativity)

=$10-($3.4*(3)+$0.4*(5)+$0*(1)+

$0*(0)) = -$2.2 <0


X*=(320,360,0)

42

(p.85)
Other Post - Optimality
Changes
(Changes in the left - hand side
coefficients.)

No

Yes

Yes

No

43

Excel Solver

Galaxy.xls
\ (Solver) .

Set Target cell $D$6


Equal To:
By Changing cells
$B$4:$C$4

Excel .

$D$7:$D$10

$F$7:$F$10
44

Excel Solver
Galaxy.xls
.

Set Target cell $D$6


Equal To:
By Changing cells
$B$4:$C$4

$D$7:$D$10<=$F$7:$F$10
/Option

.
45

Excel Solver
Galaxy.xls
Solve

Set Target cell


Equal To:

$D
$6

By Changing
$B$4:$C
cells
$4
$D$7:$D$10<=$F$7:$F$10

46

Excel Solver
GALAXY INDUSTRIES
Dozens
Profit
Plastic
Prod. Time
Total
Mix

Space Rays
320
8
2
3
1
1

Zappers
360
5
1
4
1
-1

Total
4360
1000
2400
680
-40

Limit
<=
<=
<=
<=

1000
2400
700
350

47

Excel Solver
GALAXY INDUSTRIES
Dozens
Profit
Plastic
Prod. Time
Total
Mix

Space Rays
320
8
2
3
1
1

Zappers
360
5
1
4
1
-1

Total
4360
1000
2400
680
-40

Limit
<=
<=
<=
<=

1000
2400
700
350

Solver

48

Excel Solver
Answer Report
Microsoft Excel 9.0 Answer Report
Worksheet: [Galaxy.xls]Galaxy
Report Created: 11/12/2001 8:02:06 PM

Target Cell (Max)


Cell
Name
$D$6 Profit Total

Original Value
4360

Final Value
4360

Adjustable Cells
Cell
Name
Original Value
$B$4 Dozens Space Rays
320
$C$4 Dozens Zappers
360

Final Value
320
360

Constraints
Cell
Name
$D$7 Plastic Total
$D$8 Prod. Time Total
$D$9 Total Total
$D$10 Mix Total

Cell Value
1000
2400
680
-40

Formula
$D$7<=$F$7
$D$8<=$F$8
$D$9<=$F$9
$D$10<=$F$10

Status
Slack
Binding
0
Binding
0
Not Binding
20
Not Binding
390

49

Excel Solver
Sensitivity Report
Microsoft Excel Sensitivity Report
Worksheet: [Galaxy.xls]Sheet1
Report Created:

Adjustable Cells
Final Reduced Objective Allowable Allowable
Cell
Name
Value
Cost
Coefficient Increase
Decrease
$B$4 Dozens Space Rays
320
0
8
2
4.25
$C$4 Dozens Zappers
360
0
5 5.666666667
1
Constraints
Cell
$D$7
$D$8
$D$9
$D$10

Name
Plastic Total
Prod. Time Total
Total Total
Mix Total

Final Shadow Constraint Allowable Allowable


Value
Price
R.H. Side
Increase
Decrease
1000
3.4
1000
100
400
2400
0.4
2400
100
650
680
0
700
1E+30
20
-40
0
350
1E+30
390

50


(Infeasibility):
(p.96)

(Unboundness):

( ) ( )
(p.98)

(Alternate solution):
(p.98)
51

Infeasible
Model
No point, simultaneously,
lies both above line1
2

below lines

and

2 and3

52

Solver

Solver

53


Unbounded
solution

M
ax

im

iz
e

54

Solver
Solver
Set Cell

55

Solver

Solver
LP Xj
allowable increase or
allowable decrease 0.
Solver
(p.99)
Xj
Allowable increase =
0,
Allowable decrease =

56

Solver

Objective function = Current


optimal value.
If Allowable increase = 0, change the
objective to Maximize Xj
If Allowable decrease = 0, change the
objective to Minimize Xj

Excel

57

LP


(Simplex Method)

CD3)

(Interior Point Method)


(Cutting Plane Method)
(Branch and Bound Point Method)
( CD3)
58

You might also like