Professional Documents
Culture Documents
Books 3337 0 0 (1361-1400)
Books 3337 0 0 (1361-1400)
See Figure 13. Sometimes we want to use a triangular random variable, but we are not
sure of the absolute best and worst possibilities. We may believe that there is a 10%
chance that market share will be less than or equal to 30%, that the most likely share is
40%, and that there is a 10% chance that share will exceed 75%. The RISKTRIGEN func-
tion is used in this situation. The formula
RISKTRIGEN(.3,.4,.75,10,90)
would be appropriate for this situation. Then @Risk draws a triangle that yields
■ A 10% chance that market share is less than or equal to 30%. This requires a
worst possible market share of around 20%.
■ A most likely market share of 40%.
■ A 10% chance that market share is greater than or equal to 75%. This requires a
best possible market share of around 95%.
Again, the probability of a market share between 20% and 50% is just the area under the
triangle between 20% and 50%.
FIGURE 13
Suppose that a market share of between 0 and 60% is possible, and a 45% share is
most likely. There are five market-share levels for which we feel comfortable about com-
paring relative likelihood. (See Table 1.) Thus, a market share of 45% is 8 times as likely
as 10%; 20% and 55% are equally likely; etc. Note that this distribution cannot be trian-
gular, because then 20% would be (20/45) as likely as peak of 45%, and 20% would be
.75 as likely as 45%. To model this, enter the formula
RISKGENERAL(0,60,{10,20,45,50,55},{1,6,8,7,6})
The syntax of RISKGENERAL is as follows:
■ Begin with the smallest and largest possible values.
■ Then enclose in {} the numbers for which you feel you can compare relative
likelihoods.
■ Finally, enclose in {} the relative likelihoods of the numbers you have previously
listed.
Running this in @Risk yields the output shown in Figure 15. Note that 20 is 6/8 likely as
45; 10 is 1/8 as likely as 45; 50 is 7/8 as likely as 45; 55 is 6/8 as likely as 45; etc. In be-
TA B L E 1
Market Share Relative Likelihood
10% 1
20% 6
45% 8
50% 7
55% 6
tween the given points, the density function changes at a linear rate. Thus, 30 would have
a likelihood of
(30 20)*(8 -6)
6 6.8
(45 20)
Modeling Correlations
Suppose we have three normal random variables, each having mean 0 and standard devi-
ation 1, correlated as follows:
■ Variable 1 and variable 2 have .7 correlation.
■ Variable 1 and variable 3 have a .8 correlation.
■ Variable 2 and variable 3 have a .75 correlation.
To model this correlation structure, we use the RISKCORRMAT command. Simply en-
ter your correlation matrix somewhere in the worksheet. In Figure 16, we chose C27:E29.
B C D E F G H
25
26
27 1 0.7 0.8
28 0.7 1 0.75
29 0.8 0.75 1
30
31 1 Variable 1 1.793028 risknormal(0,1,riskcorrmat(c27:e29,1))
32 2 Variable 2 -0.449129 risknormal(0,1,riskcorrmat(c27:e29,2))
33 3 Variable 3 -0.521328 risknormal(0,1,riskcorrmat(c27:e29,3))
FIGURE 16
FIGURE 17
Cases
Jeffrey B. Goldberg
UNIVERSITY OF ARIZONA
TA B L E 2
Product-Resource Constraints: Glendale
Resources
Labor Packing Advertising
Products (Minutes/Unit) (Minutes/Unit) ($/Unit)
TA B L E 3
Product-Resource Constraints: Tucson
Resources
Labor Packing Advertising
Products (Minutes/Unit) (Minutes/Unit) ($/Unit)
TA B L E 5
Minimum and Maximum Product Demand per Month
Store (or Plant)
Products Chandler Glendale Tucson
Vision is a large company that produces video- Total available 3,000 6,000 —
capturing devices for military applications such as
missiles, long-range cameras, and aerial drones. Four
different types of cameras (differing mainly by lens pounds of material is available for the entire system
type) are made in the three plants in the system. Each during the planning period.
plant can produce any of the four camera types, Transport has 3 major customers (RAYco, HONco,
although each plant has its own individual constraints and MMco) for its products. The maximum sales for
and unit costs. These constraints cover labor and ma- each customer–product pair is given in Table 11. Prod-
chining restrictions, and the specific values are given uct sales prices are given in Table 12, and the ship-
in Tables 8–10. Note that even though the products ping costs from each plant to each customer are de-
are identical in the three plants, different production tailed in Table 13. Table 14 contains the production
processes are used and thus the products use different costs for each product–plant pair.
amounts of resources in different plants. The corpo- All shipping from plants 1 and 2 that goes to
ration controls the material that goes into the lenses; RAYco or HONco must go through a special inspec-
the material requirements for each product are given tion. These units are sent to a central site, inspected,
in the last column of Tables 8–10. A total of 3,500 and then sent to their destination. The capacity of this
special inspection site is 1,500 pieces.
Your job is to determine a recommendation for the
TA B L E 8
company. A recommendation must include a plan for
Product-Resource Constraints: Plant 1 production and shipping as well as the cost and rev-
Resources enue generated from each plant. In addition, you
Labor Machine Material should address the following potential issues in your
Products (Hours/Unit) (Hours/Unit) (Lb./Unit)
recommendation:
Small 3 8 1.0
■ If you could get more material, how much would
Medium 3 8.5 1.1
you like? How would you use it? What would you
Large 4 9 1.2
be willing to pay?
Precision 4 9 1.3
■ If you could get more inspection capacity, how
Total available 6,000 10,000 —
much would you like? How would you use it? What
would you be willing to pay?
1 AFC 1 28 FSM 30
2 AFC 2 27 FSM 31
3 LSM 1 26 32
4 LSM 2 25
5 OCR 1 24 Carrier
parking lot
OCR 2 Node 33
6 23
7 OCR 3 22
8 BCS 1 21 Staging
interface
shipping to
9 10 11 12 13 20 other GMFs
and carrier
stations
Node 34
BCS DBCS DBCS DBCS
2 1 2 3
16
14 15 16 17 18
TA B L E 17
Salary and Skills Required for Each Job Classification
Person Salary ($) Skills 1–20
Senior Manager 1 1 1 1 0 1 1 0 1 0 1 1 0 1 0 1 1 1 0 0 0
Project Manager 0 1 1 1 0 1 1 1 1 0 1 0 1 0 0 1 1 1 1 0 1
Professional 0 0 0 0 1 1 0 0 0 1 0 0 1 0 1 0 0 1 1 0 1
Sales 0 0 0 0 1 1 0 0 0 1 0 0 1 0 1 0 0 1 1 1 1
Technician 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1
Administrative 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 1 1 0
Assistant
Program 1 500 1 1 0 1 0 0 0 0 0 0 1 1 1 1 1 0 1 0 0 0
Program 2 300 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0
Program 3 500 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 1 0 0 0 1
Program 4 575 0 1 0 1 0 0 0 0 1 1 0 0 1 0 0 0 1 0 0 0
Program 5 800 0 1 1 1 0 0 0 0 0 1 0 0 1 0 0 0 1 0 0 0
Program 6 400 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 1
Program 7 200 0 0 0 0 1 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1
Program 8 1,000 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 0 0 1 0
Program 9 200 1 1 1 1 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0
Program 10 500 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0
Program 11 700 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0
Program 12 600 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0
Program 13 400 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1 1 1 0
Program 14 900 1 0 0 0 1 1 0 0 0 0 0 0 0 0 0 1 0 1 0 0
Program 15 700 1 0 1 0 1 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0
Program Skills 21–41
Program 1 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
Program 2 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0
Program 3 1 1 1 1 1 1 1 0 1 0 1 0 0 0 0 1 0 0 0 0 0
Program 4 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 1
Program 5 0 0 0 1 0 1 0 0 0 0 0 0 1 0 0 0 0 0 1 1 0
Program 6 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 1 1 0 0 0 0
Program 7 0 1 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0
Program 8 1 1 1 1 0 0 1 0 1 1 0 0 0 0 0 0 1 0 0 0 0
Program 9 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0
Program 10 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 1 1 0 0 0
Program 11 0 1 1 0 0 0 1 0 1 1 0 1 0 0 1 1 0 0 0 0 0
Program 12 0 0 0 0 0 0 0 1 0 1 1 0 0 0 0 0 0 1 1 1 0
Program 13 1 0 1 0 1 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0
Program 14 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 1
Program 15 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 1
■ Employees are divided into classes. In each class, ■ Training is equally effective for all people, thus we
we have estimates for (1) the number of employ- are concerned with which programs to offer and
ees, (2) the employee hourly wage, (3) the number which employees in each class to assign to each
of employees that require each particular skill, and program. If we know the quality of the training for
(4) the maximum time available for training em- individual skills for individual classes, then we can
ployees in each class during the study period. relax this assumption.
■ There is a list of training programs. For each pro-
gram, we assume we have (1) the set of skills Potential Corporate Setting
taught, (2) the cost, (3) the development time,
(4) the completion time for an employee, and Your job is to develop models to aid businesses and
(5) the maximum number of employees who can corporations in determining the appropriate training
participate per decision cycle. programs to use. The type of model and issues often
depend on the size of the corporation and the potential
Yuma, AZ 125,000 10 5 16
Fresno, CA 130,000 12 8 11
Tucson, AZ 140,000 9 10 15
Pomona, CA 160,000 11 7 14
Santa Fe, NM 150,000 8 14 10
Flagstaff, AZ 170,000 10 12 11
Las Vegas, NV 155,000 13 12 9
St. George, UT 115,000 14 15 8
TA B L E 21
Demand Information
Demand
Company Location Regular Green Onion Party Mix
FIGURE 4
Map of the City (17 miles by 11 miles)
X
2 X
X
1
3 4
X
5
X
X
6
X X
8 10
7 X X 9
X marks the spot of the existing sites and the possible sites in each zone
TA B L E 23
Demand Information per Year
Zone Fire Calls False Alarms Medical Calls Base Cost ($/Year)
System Design: Project Management† This is only the quality for a task. It is unclear how
one might quantify the quality of the project.
Your job is to determine how we should proceed
System Design (SD) is a small corporation that con- with our tasks. Your analysis should answer some of
tracts to manage systems and industrial engineering the following questions:
projects. In this case it must manage the design and
construction of a power plant’s data-processing and ■ What tasks are critical to project completion? What
data-collection system. SD’s role in the project is to tasks will you expedite?
■ How are you measuring system quality, and how
†
does your recommendation measure up relative to
This material is expanded from a homework problem in Applied
that objective?
Mathematical Programming by Wayne Winston.
TA B L E 25
Expediting Costs ($1,000) and Limits
Duration (Days)
Task No. Current tj Minimum dj Cost to Decrease tj by x Days
2 100 70 1.5x1.8
6 40 20 2x2.0
7 35 20 1x2.0
10 70 40 1.8x1.9
14 80 60 1.9x1.6
15 130 120 0.95x2.7
19 125 80 0.9x2.9
TA B L E 27
Item Cost and Base Size
Item Cost ($) Base Unit Base Unit Weight (Lb.)
Ace bandages 1 2 4 12 6
Band-Aids 0 2 4 4 4
Flares 2 1 1 0 2
Blankets 1 1 4 2 3
Adhesive tape 2 2 3 6 4
Cold packs 2 2 3 6 3
Sunburn cream 1 2 4 4 5
Antiseptic cream 1 2 3 2 4
Acetaminophen caplets 1 2 4 6 6
Rubber gloves 1 1 2 10 5
FIGURE 5
Modular Assembly
Modules
Band-Aids Gloves
Flares Blankets
1 Linearly dependent.
SECTION 2.1
2 Linearly independent.
1 2 3
1 a A 4 5 6 SECTION 2.5
7 8 9
12 12 12
1
1 2 3
3 6 9 2 A
12 12 12
b 3A 12 15 18
21 24 27 3 A1 does not exist.
1
8 a 1B .
c A 2B is undefined. 100
1 4 7
SECTION 2.6
d AT 2 5 8
3 6 9 2 30.
0 1
2
1
e BT REVIEW PROBLEMS
1 2
1 Infinite number of solutions of the form x1 k 1,
4 6 x2 3 k, x3 k.
f AB 10 15
T 0.20 T
Ut1 0.75 0 Ut
3
16 24 t1 0.90 t
g BA is undefined. 4 x1 0, x2 1.
y1 0.50 0 0.10 x1 13 Linearly independent.
2 y2 0.30 0.70 0.30 x2 14 Linearly dependent.
y3 0.20 0.30 0.60 x3
15 a Only if a, b, c, and d are all nonzero will rank A
4. Thus, A1 exists if and only if all of a, b, c, and d
SECTION 2.2 are nonzero.
b Applying the Gauss-Jordan method, we find if a, b,
1 1 4 1 1 4 c, and d are all nonzero,
x
1 2 1 1 6 or 2 1 6
x 1 0 0 0
1 3 2 8 1 3 8 a
1
0 0 0
A1 b
0 0 1 0
SECTION 2.3 c
0 0 0 1d
1 No solution. 18 4.
2 Infinite number of solutions of the form x1 2 2k,
x2 2 k, x3 k.
3 x1 2, x2 1.
Chapter 3 s.t. 8(x1 x4 x5 x6 x7) 4(x8 x11 x12
x13 x14) 88 (Sunday)
SECTION 3.1
s.t. 8(x1 x2 x5 x6 x7) 4(x8 x9 x12
1 max z 30x1 100x2 x13 x14) 136 (Monday)
1 s.t. x1 x2 7 (Land constraint) s.t. 8(x1 x2 x3 x6 x7) 4(x8 x9 x10
4x1 10x2 40 (Labor constraint) x13 x14) 104 (Tuesday)
1 s.t. 10x1 + 10 x2 30 (Government constraint) s.t. 8(x1 x2 x3 x4 x7) 4(x8 x9 x10
x1, x2 0 x11 x14) 120 (Wednesday)
2 No, the government constraint is not satisfied. s.t. 8(x1 x2 x3 x4 x5) 4(x8 x9 x10
b No, the labor constraint is not satisfied.
x11 x12) 152 (Thursday)
c No, x2 0 is not satisfied.
s.t. 8(x2 x3 x4 x5 x6) 4(x9 x10 x11
SECTION 3.2 x12 x13) 112 (Friday)
4 a We want to make x1 larger and x2 smaller, so we 20(x8 x9 x10 x11 x12 x13 x14)
move down and to the right. 0.25(136 104 120 152 112 128
b We want to make x1 smaller and x2 larger, so we
88)
move up and to the left.
b We want to make both x1 and x2 smaller, so we move (The last constraint ensures that part-time labor
down and to the left. will fulfill at most 25% of all labor requirements)
All variables 0
SECTION 3.3 3 Let x1 number of employees who start work on Sunday
1 No feasible solution. and work five days, x2 number of employees who start
work on Monday and work five days, . . . , x7 number of
2 Alternative optimal solutions. employees who start work on Saturday and work five days.
3 Unbounded LP. Also let o1 number of employees who start work on
Sunday and work six days, . . . , o7 number of employees
who start work on Saturday and work six days. Then the
SECTION 3.4 appropriate LP is
1 For i 1, 2, 3, let xi tons of processed factory i min z 250(x1 x2 x7)
waste. Then the appropriate LP is min z + 312(o1 o2 o7)
min z 15x1 10x2 20x3 s.t. x1 x4 x5 x6 x7 o1 o3 o4
s.t. 0.10x1 0.20x2 0.40x3 30 (Pollutant 1) o5 o6 o7 11 (Sunday)
s.t. 0.45x1 0.25x2 0.30x3 40 (Pollutant 2) s.t. x1 x2 x5 x6 x7 o1 o2 o4
x1, x2, x3 0 o5 o6 o7 17 (Monday)
It is doubtful that the processing cost is proportional to the s.t. x1 x2 x3 x6 x7 o1 o2 o3
amount of waste processed. For example, processing 10
tons of waste is probably not 10 times as costly as processing o5 o6 o7 13 (Tuesday)
1 ton of waste. The Divisibility and Certainty Assumptions s.t. x1 x2 x3 x4 x7 o1 o2 o3
seem reasonable.
o4 o6 o7 15 (Wednesday)
SECTION 3.5 s.t. x1 x2 x3 x4 x5 o1 o2 o3
2 Let x1 number of chocolate cakes baked and x2 2(M21 NM21) (M22 NM22)
number of vanilla cakes baked. Then we must solve (M31 NM31) (M32 NM32)
max z x1 12 x2 s.t. M11 M12 50
1 2
s.t. x1
3
x2
3
8 M21 M22 50
4x1 x2 30 M31 M32 100
x1, x2 0 NM11 NM12 200
$69 36 66
The optimal solution is z ,
7
x1 ,
7
x2 .
7 NM21 NM22 250
8 Let x1 acres of farm 1 devoted to corn, x2 acres of NM31 NM32 150
farm 1 devoted to wheat, x3 acres of farm 2 devoted to
corn, x4 acres of farm 2 devoted to wheat. Then a correct
formulation is
(Mediocre chips)
1 x1 4,999, x2 5,000 has z 10,000.
s.t. 400x1 300x2 250x3 s4 s4 28,000
SECTION 4.10 (Budget constraint)
1 a Both very small numbers (for example, 0.000003) All variables 0
and large numbers (for example, 3,000,000) appear in
the problem. REVIEW PROBLEMS
b Let x1 units of product i produced (in millions).
If we measure our profit in millions of dollars, the LP 4 Unbounded LP.
becomes 5 z 6, x1 0, x2 3.
1 max 6x1 4x2 3x3 6 Infeasible LP.
1 s.t. 4x1 3x2 2x3 3 (Million labor hours) 8 z 12, x1 x2 2.
1 s.t. 3x1 2x2 x3 2 (lb of pollution) 10 Four types of furniture.
1 s.t. 3x1, x2, x3 0 15 z 17
, x2 32, x4 12.
2
17 a c 0 and b 0.
SECTION 4.11
b b 0 and c 0. Also need a2 > 0 and/or a3 > 0
1 z 16, x1 x2 2. The point where all three constraints to ensure that when x1 is pivoted in, a feasible solution
are binding (x1 x2 2) corresponds to the following results. If only a3 > 0, then we need b to be strictly
three sets of basic variables: positive.
20 Let ct net number of drivers hired at the beginning B1, B2, B3, B4 2
of the year t. Then ct ht ft, where ht number of drivers All variables 0
hired at beginning of year t, and ft number of drivers fired
at beginning of year t. Also let dt number of drivers after
drivers have been hired or fired at beginning of year t. Then Chapter 5
a correct formulation is (cost in thousands of dollars)
min z 10(d1 d2 d3 d4 d5) SECTION 5.1
2(f1 f2 f3 f4 f5) 1 Decision variables remain the same. New z-value is
4(h1 h2 h3 h4 h5) $210.
s.t. d1 50 h1 f1 4 a 530 c1 350.
c 4,000,000 HIW 84,000,000; x1 3.6 0.15,
s.t. d2 d1 h2 f2
x2 1.4 0.025.
s.t. d3 d2 h3 f3 f $310,000.
s.t. d4 d3 h4 f4
s.t. d5 d4 h5 f5 SECTION 5.2
s.t. R3 R2 B3 S3
s.t. R4 R3 B4 S4
I
1 I 1 C1 600
I
2 I2 I
1 I 1 C2 800
I
3 I 3 I
2 I2 C3 500
I
4 I4 I
3 I 3 C4 400
s.t. R4 2
s.t. C1 200R1
s.t. C2 200R2
1 a $1,046,667.
b Yes.
c $33.33.
b $333.33.
7 a Decision variables remain unchanged. New z-value
is $1,815,000.
b Pay $0 for an additional 100 board ft of lumber. Pay
$1,350 for an additional 100 hours of labor.
c $1,310,000.
d $1,665,000.
Chapter 6
SECTION 6.1
1 z 4s1 5s2 28
x1 s1 s2 6
x2 s1 2s2 10
SECTION 6.3
SECTION 6.4
1 Yes.
2 No.
4 Yes.
Chapter 7
SECTION 7.1
SECTION 7.6
1 a
REVIEW PROBLEMS
3 Meet January demand with 30 units of January from Bi to Bj means one black student from district i goes
production. Meet February demand with 5 units of January to school in district j. The costs of M ensure that shipments
production, 10 units of February production, and 15 units of from Wi to Bj or Bi to Wj cannot occur (table on next page).
March production. Meet March demand with 20 units of
8 Optimal solution is z 1,580, x11 40, x12 10, x13
March production.
10, x22 50, x32 10, x34 30.
4 Maid 1 does the bathroom, maid 2 straightens up, maid
13 Optimal solution is z 98, x13 5, x21 3, x24 7,
3 does the kitchen, maid 4 gets the day off, and maid 5
x32 3, x33 7, x34 5.
vacuums.
25 Sell painting 1 to customer 1, painting 2 to customer
7 Shipping 1 unit from Wi to Wj means one white student
2, painting 3 to customer 3, and painting 4 to customer 4.
from district i goes to school in district j. Shipping 1 unit
SECTION 8.2
M large number to prevent shipping a unit through a 1 s.t. x0 xso,1 xso,2 (Node so)
nonexistent arc.
1 s.t. xso,1 x13 x12 (Node 1)
5 Replace the car at times 2, 4, and 6. Total cost $14,400.
1 s.t. x12 x32 xso,2 x24 (Node 2)
FIGURE 5 d s.t. x2 x1 3
d s.t. x3 x2 6
d s.t. x4 x2 2
d s.t. x5 x4 3
d s.t. x5 x3 1
d s.t. x6 x5 1.5
d s.t. x8 x6 2
d s.t. x8 x7 3(x7 x3 2)
d s.t. x8 x2 3
d s.t. x8 x3 5
All variables urs
5 a See Figure 8. A–B–E–F–G and A–B–C–G are critical
paths. Duration of project is 26 days.
FIGURE 8
7 See Figure 6. If maximum flow 4, then all jobs can
be completed.
FIGURE 6
A 0 0
B 0 0
C 0 0
D 8 8
E 0 0
F 0 0
G 0 0