Solving Examples of Linear Program-

ming Models

Prof. Yong Won SEO

([email protected])

College of Business Administration, CAU

Chapter Topics

 A Product Mix Example

 A Diet Example
 An Investment Example
 A Marketing Example
 A Transportation Example
 A Blend Example
 A Multiperiod Scheduling Example
 A Data Envelopment Analysis Example
Product Mix Example
Product Mix Example

Four-product T-shirt/sweatshirt manufacturing company.

■ Must complete production within 72 hours
■ Truck capacity = 1,200 standard sized boxes.
■ Standard size box holds 12 T-shirts.
■ One-dozen sweatshirts box is three times size of standard box.
■ $25,000 available for a production run.
■ 500 dozen blank T-shirts, and 500 dozen blank sweatshirts in
stock.
■ How many dozens (boxes) of each type of shirt to produce?
 Product Mix Example
Resource requirements for the product mix example.
Processing Cost Profit
Time (hr) ($) ($)
Per dozen per dozen per dozen
Sweatshirt - F 0.10 $36 $90
Sweatshirt – B/ F 0.25 48 125
T-shirt - F 0.08 25 45
T-shirt - B/ F 0.21 35 65
Product Mix Example
Decision Variables:
x1 = sweatshirts, front printing
x2 = sweatshirts, back and front printing
x3 = T-shirts, front printing
x4 = T-shirts, back and front printing
Objective Function:
Maximize Z = $90x1 + $125x2 + $45x3 + $65x4
Model Constraints:
0.10x1 + 0.25x2+ 0.08x3 + 0.21x4  72 hr
3x1 + 3x2 + x3 + x4  1,200 boxes
$36x1 + $48x2 + $25x3 + $35x4  $25,000
x1 + x2  500 dozen sweatshirts
x3 + x4  500 dozen T-shirts
Product Mix Example

• Model solution is:

• x1=175.56 boxes of front-only sweatshirts
• x2=57.78 boxes of front and back sweatshirts
• x3 = 500 boxes of front-only t-shirts
• Z=$45,522.22 profit
Product Mix Example: Sensitivity

• Additional 1 hour of processing time means

$233.33 increased profit
 Diet Example
Breakfast Food Fat Cholesterol Iron Calcium Protein Fiber Cost
Cal (g) (mg) (mg) (mg) (g) (g) ($)
1. Bran cereal (cup) 90 0 0 6 20 3 5 0.18
2. Dry cereal (cup) 110 2 0 4 48 4 2 0.22
3. Oatmeal (cup) 100 2 0 2 12 5 3 0.10
4. Oat bran (cup) 90 2 0 3 8 6 4 0.12
5. Egg 75 5 270 1 30 7 0 0.10
6. Bacon (slice) 35 3 8 0 0 2 0 0.09
7. Orange 65 0 0 1 52 1 1 0.40
8. Milk-2% (cup) 100 4 12 0 250 9 0 0.16
9. Orange juice (cup) 120 0 0 0 3 1 0 0.50
10. Wheat toast (slice) 65 1 0 1 26 3 3 0.07

Breakfast to include at least 420 calories, 5 milligrams of iron,

400 milligrams of calcium, 20 grams of protein, 12 grams of
fiber, and must have no more than 20 grams of fat and 30
milligrams of cholesterol.
Diet Example

x1 = cups of bran cereal

x2 = cups of dry cereal
x3 = cups of oatmeal
x4 = cups of oat bran
x5 = eggs
x6 = slices of bacon
x7 = oranges
x8 = cups of milk
x9 = cups of orange juice
x10 = slices of wheat toast
 Diet Example
Minimize Z = 0.18x1 + 0.22x2 + 0.10x3 + 0.12x4 + 0.10x5 + 0.09x6
+ 0.40x7 + 0.16x8 + 0.50x9 + 0.07x10
subject to:
90x1 + 110x2 + 100x3 + 90x4 + 75x5 + 35x6 + 65x7
+ 100x8 + 120x9 + 65x10  420 calories
2x2 + 2x3 + 2x4 + 5x5 + 3x6 + 4x8 + x10  20 g fat
270x5 + 8x6 + 12x8  30 mg cholesterol
6x1 + 4x2 + 2x3 + 3x4+ x5 + x7 + x10  5 mg iron
20x1 + 48x2 + 12x3 + 8x4+ 30x5 + 52x7 + 250x8
+ 3x9 + 26x10  400 mg of calcium
3x1 + 4x2 + 5x3 + 6x4 + 7x5 + 2x6 + x7
+ 9x8+ x9 + 3x10  20 g protein
5x1 + 2x2 + 3x3 + 4x4+ x7 + 3x10  12
Diet Example : Solution

• X3=1.025
• X8=1.241
• X10=2.975

• Z=$0.509 per meal

Diet Problem History

• The first problem tested on by Dantzig

• Originally formulated in 1945 by Nobel economist
George Stigler
– An important military issue during WW II.
• Consisted of 77 unknowns and 9 equations
– 9 clerks using hand-operated desk calculators
– 120 MD to obtain the optimal Simplex solution
• Solution
– $39.69 per year using wheat flour, cabbage, dried navy beans
(1939 prices)
– Simplex solution was better than Stiegler’s (using his own
method), by 24 cents
 Investment Example
An investor has $70,000 to divide among several instruments.
Municipal bonds have an 8.5% return, CD’s a 5% return, t-bills a 6.5%
return, and growth stock 13%.

The following guidelines have been established:

1.No more than 20% in municipal bonds
2.Investment in CDs should not exceed the other three alternatives
3.At least 30% invested in t-bills and CDs
4.More should be invested in CDs and t-bills than in municipal bonds
and growth stocks by a ratio of 1.2 to 1
5.All $70,000 should be invested.
 Investment Example
Maximize Z = $0.085x1 + 0.05x2 + 0.065 x3+ 0.130x4
subject to:
x1  $14,000
x2 - x1 - x3- x4  0
x2 + x3  $21,000
-1.2x1 + x2 + x3 - 1.2 x4  0
x1 + x2 + x3 + x4 = $70,000
x1, x2, x3, x4  0
where
x1 = amount ($) invested in municipal bonds
x2 = amount ($) invested in CD(certificates of deposit)
x3 = amount ($) invested in t-bills(treasury bills)
x4 = amount ($) invested in growth stocks
Investment Example
Objective function, Z,
for total return

First
guideline,
=D6*B13

Total investment requirement,

=D10*B13+E10*B14+F10*B15+G10*B16
Investment Example: Sensitivity

• Additional $1 investment yiels $0.095 (9.5%) increase in

profit, with no upper bound

Shadow price for the

amount available to
invest
Investment Example: variation

• If the following guideline is removed, how to model?

5. All $70,000 should be invested.

[Case]
Investment Application of GE Asset Mgt
Marketing Example
Exposure
(people/ad or Cost
commercial)
Television Commercial 20,000 $15,000

Radio Commercial 2,000 6,000

Newspaper Ad 9,000 4,000

 Budget limit $100,000

 Television time available for 4 commercials
 Radio time available for 10 commercials
 Newspaper space available for 7 ads
 Resources for no more than 15 commercials and/or ads
Marketing Example

Maximize Z = 20,000x1 + 12,000x2 + 9,000x3

subject to:
15,000x1 + 6,000x 2+ 4,000x3  100,000
x1  4
x2  10
x3  7
x1 + x2 + x3  15
x1, x2, x3  0
where
x1 = number of television commercials
x2 = number of radio commercials
x3 = number of newspaper ads
Marketing Example
• TV: 1.818, Radio: 10, Newspaper ads: 3.182 (????)
– Total Exposures 185000
– Round down?
Marketing Example: Forcing Integer

Decision
variables

Click on “int” for

integer.

Exhibit 4.12
Marketing Example: Integer Solution

Better solution—17,000
Integer solution more total exposures—
than rounded-down
solution

Exhibit 4.14
Transportation Example
Warehouse supply of Retail store demand
Television Sets: for television sets:
1 - Cincinnati 300 A - New York 150
2 - Atlanta 200 B - Dallas 250
3 - Pittsburgh 200 C - Detroit 200
Total 700 Total 600
Unit Shipping Costs:

To Store
From Warehouse
A B C
1 $16 $18 $11
2 14 12 13
3 13 15 17
Transportation Example

Minimize Z = $16x1A + 18x1B + 11x1C + 14x2A + 12x2B + 13x2C +

13x3A + 15x3B + 17x3C
subject to:
x1A + x1B+ x1C  300
x2A+ x2B + x2C  200
x3A+ x3B + x3C  200
x1A + x2A + x3A = 150
x1B + x2B + x3B = 250
x1C + x2C + x3C = 200
All xij  0
Transportation Example: Solution

=C5+D5+E5
=C5+C6+C7
Blend Example

■ Determine the optimal mix of the three components in each grade

of motor oil that will maximize profit. Company wants to
produce at least 3,000 barrels of each grade of motor oil.
Blend Example

Maximum Barrels
Component Cost/barrel
Available/day
1 4,500 $12
2 2,700 10
3 3,500 14

Grade Component Specifications Selling Price ($/bbl)

At least 50% of 1
Super $23
Not more than 30% of 2
At least 40% of 1
Premium
Not more than 25% of 3 20
At least 60% of 1
Extra 18
At least 10% of 2
Blend Example

• Decision variables: The quantity of each of the three compo-

nents used in each grade of gasoline (9 decision variables); xij =
barrels of component i used in motor oil grade j per day, where i
= 1, 2, 3 and j = s (super), p (premium), and e (extra).
 Blend Example
Maximize Z = 11x1s + 13x2s + 9x3s + 8x1p + 10x2p + 6x3p + 6x1e
+ 8x2e + 4x3e
subject to:
x1s + x1p + x1e  4,500 bbl.
x2s + x2p + x2e  2,700 bbl.
x3s + x3p + x3e  3,500 bbl.
0.50x1s - 0.50x2s - 0.50x3s  0
0.70x2s - 0.30x1s - 0.30x3s  0
0.60x1p - 0.40x2p - 0.40x3p  0
0.75x3p - 0.25x1p - 0.25x2p  0
0.40x1e- 0.60x2e- - 0.60x3e  0
0.90x2e - 0.10x1e - 0.10x3e  0
x1s + x2s + x3s  3,000 bbl.
x1p+ x2p + x3p  3,000 bbl. all xij  0
x1e+ x2e + x3e  3,000 bbl.
 Blend Example: Solution

=B7+B10+B13

=0.5*B7-0.5*B8-
Decision variables— =B7+B8+B9 0.5*B9
B7:B15
Blend Example: Sensitivity
• Component 1 seems the most critical resource
– 1 barrel increase means $20 additional profit, up to 6200 barrels
[Case]
Blending Problem of Petroleum Industry
Multi-period Scheduling Example

Production Capacity: 160 computers per week

50 more computers with overtime
Assembly Costs: $190 per computer regular time;
$260 per computer overtime
Inventory Holding Cost: $10/computer per week
Order schedule:
Week Computer Orders
1 105
2 170
3 230
4 180
5 150
6 250
Multi-period Scheduling Example

Decision Variables:
rj = regular production of computers in week j
(j = 1, 2, …, 6)
oj = overtime production of computers in week j
(j = 1, 2, …, 6)
ij = extra computers carried over as inventory in week j
(j = 1, 2, …, 5)
Multi-period Scheduling Example
Model summary:
Minimize Z = $190(r1 + r2 + r3 + r4 + r5 + r6) +
$260(o1+o2+o3 +o4+o5+o6) + 10(i1 + i2 + i3 + i4 + i5)
subject to:
rj  160 computers in week j ( j = 1, 2, 3, 4, 5, 6)
oj  50 computers in week j ( j = 1, 2, 3, 4, 5, 6)
r1 + o1 - i1 = 105 week 1
r2 + o2 + i1 - i2 = 170 week 2
r3 + o3 + i2 - i3 = 230 week 3
r4 + o4 + i3 - i4 = 180 week 4
r5 + o5 + i4 - i5 = 150 week 5
r6 + o6 + i5 = 250 week 6
rj, oj, ij  0
 Multi-period Scheduling: Solution

G7-H7

B7+D7+I6; regular
Decision variables for Decision variables for production + overtime
regular production – overtime production production + inventory
B6:B11 – D6:D11 from previous week

Exhibit 4.20
[Case] Employee Scheduling Problem
DEA (Data Envelopment Analysis)

DEA compares a number of decision making units(DMU) of the

same type based on their inputs (resources) and outputs.
The result indicates if a particular unit is less productive, or
efficient, than other units.
[Case] DEA to compare ports’ efficiency
 Comparing the efficiency

• Which department is more efficient?

Department Input: Output:
Labor Cost($) Sales(# of contracts)
A 2000 1500
B 1500 1100

• Comparing Sales/Labor cost

De- Input: Output: Sales/Labor

part- Labor Cost($) Sales(# of con- Cost
ment tracts)
A 2000 1500 0.75
B 1500 1100 0.73

42
 Comparing the efficiency is not easy

• Which department is more efficient?

Depart- Input 1: Input 2: Output:
ment Labor Cost($) Office Space(ft2) Sales(# of contracts)
A 2000 10000 1500
B 1500 6900 1100

• Measuring the efficiency(or performance) of different

organizations is difficult because organizations have
multiple input measures (number of workers, cost of la-
bor, cost of machine operations, pay scale for employ-
ees, and cost of advertising) and multiple output mea-
sures (profit, sales, and market share).
43
 Calculating Efficiency

• DEA offers a variety of models that use multiple inputs

and outputs to compare the efficiency of two or more
processes.

• The ratio model is based on the following definition of

efficiency:

Weighted Sum of Outputs

Efficiency =
Weighted Sum of Inputs

44
 Calculating Efficiency (cont’d)

• Suppose we have the following input and output data:

Department Labor Cost Sales Customer Satisfaction
(input) (output 1) (output 2)
A 10 10 10
B 15 30 12
C 12 36 6
D 22 25 16

• Defining efficiency based on the ratio model:

w1  Sales  w2  CS
Efficiency 
v  Labor

• The rationale is : “Choose whatever weights that would

45 make your efficiency value maximum”
 Building an LP for Department D
25w1  16 w2 Maximize 25w1  16 w2
Maximize
22v
Subject to
Subject to
22v  1
10 w1  10 w2
1 10 w1  10 w2  10v
10v
30 w1  12 w2 30 w1  12 w2  15v
1 LP 36 w1  6 w2  12v
15v
36 w1  6 w2 25w1  16w2  22v
1
12v w1 , w2 , v  0
25w1  16 w2
1
22v
w1 , w2 , v  0
46
DEA Solution Meaning

• 4 DEA formulations can be made

– To Evaluate A, B, C, D each

• If Z=1, the school is efficient one

• If Z<1, the school is less efficient than others
 Graphical Analysis

CS/Labor Cost
1.2
A
1 Efficient frontier
B
0.8 D
0.6 C
0.4

0.2

0
0 0.5 1 1.5 2 2.5 3 3.5
Sales/Labor Cost

48
DEA Example

Elementary school comparison:

Input 1 = teacher to student ratio
Input 2 = supplementary funds/student
Input 3 = average educational level of parents

Output 1 = average reading SOL score

Output 2 = average math SOL score
Output 3 = average history SOL score
DEA Example: 4 schools’ data

Inputs Outputs

School 1 2 3 1 2 3

Alton .06 $260 11.3 86 75 71

Beeks .05 320 10.5 82 72 67

Carey .08 340 12.0 81 79 80

Delancey
.06 460 13.1 81 73 69
DEA Example

• Objective: To compare 4 schools’ efficiency

• Multiple elements in I/O : weights needed

• Evaluation can vary with weight settings

– What weight values seems fair to you?

• Rationale: “Choose whatever weights that would make

your efficiency value maximum”
DEA to evaluate Delancey
x1  81  x2  73  x3  69
Maximize
y1  0.06  y2  460  y3 13.1
Subject to
x1  86  x2  75  x3  71
1
y1  0.06  y2  260  y3 11 .3
x1  82  x2  72  x3  67
1
y1  0.05  y2  320  y3 10.5
x1  81  x2  79  x3  80
1
y1  0.08  y2  340  y3 12.0
x1  81  x2  73  x3  69
1
y1  0.06  y2  460  y3 13.1
All variables  0
DEA to evaluate Delancey – LP form

Decision Variables:
xi = weight per unit of each output where i = 1, 2, 3
yi = weight per unit of each input where i = 1, 2, 3
Model Summary:
Maximize Z = 81x1 + 73x2 + 69x3
subject to:
.06 y1 + 460y2 + 13.1y3 = 1
86x1 + 75x2 + 71x3 .06y1 + 260y2 + 11.3y3
82x1 + 72x2 + 67x3  .05y1 + 320y2 + 10.5y3
81x1 + 79x2 + 80x3  .08y1 + 340y2 + 12.0y3
81x1 + 73x2 + 69x3  .06y1 + 460y2 + 13.1y3
xi, yi  0
DEA Solution: Z=Efficiency

=E8*D12+F8*D13+G8*D14
=B5*B12+C5*B13+D5*B1
4

Value of outputs, also in cell H8

Exhibit 4.22