Professional Documents
Culture Documents
Spreadsheet Modeling and Decision Analysis A Practical Introduction To Business Analytics 7th Edition Cliff Ragsdale Solutions Manual Download
Spreadsheet Modeling and Decision Analysis A Practical Introduction To Business Analytics 7th Edition Cliff Ragsdale Solutions Manual Download
————————————————————————————————————————————
Chapter 6
Integer Linear Programming
1. The optimal solutions to an LP problem occur at extreme points of the feasible region. There are a finite
number of such extreme points for any given problem and they can be located and searched very
efficiently. The optimal solution to an ILP does not necessarily occur at an extreme point of the relaxed
feasible region, making it more difficult to locate these points.
3. Errata: The objective in this problem should be MIN rather than MAX.
See file: Prb6_3.xlsm
a. X1=6.25, X2=6.5
b. X1=8, X2=8
c. The optimal integer solution to an ILP cannot, in general, be obtained by rounding the continuous
solution.
4. a. X1 + X2 + X4 + X6 = 2
b. X2 - X3 = 0
c. 2X5 - X3 - X4 0
d. X2 + X 4 - X5 1
6. a. It would merely serve to reduce his net profit by $1,000. Since he must incur this cost regardless of
what kind of hot tub he makes there is not reason to include this cost in the formulation of the model.
b. MAX 350 X1 + 300 X2 - 900 Y1 - 800 Y2
ST 1 X1 + 1 X2 200
9 X1 + 6 X2 1,520
12 X1 + 16 X2 2,650
X1 - M Y 1 0
X2 - M Y 2 0
Chapter 6 - Integer Linear Programming : S-2
————————————————————————————————————————————
8. a. MIN X1 + X 2 + X 3 + X 4 + X 5 + X 6 + X 7
ST X2 + X3 + X4 + X5 + X6 18
X3 + X4 + X5 + X6 + X7 17
X1 + X4 + X5 + X6 + X7 16
X1 + X2 + X5 + X6 + X7 16
X1 + X2 + X3 + X6 + X7 16
X1 + X2 + X3 + X4 + X7 14
X1 + X2 + X3 + X4 + X5 19
Xi 0 & integer
X6 1
Xi 0 & integer
10. a. Sum the sales potentials & divide by 5 = 210. This would give each sales rep an equal sales potential.
b. Ui = amount by which rep i is under the target amount
Oi = amount by which rep i is under the target amount
Xij = 1 if rep i is assigned to account j; 0 otherwise
MIN: U1 + O1 + U2 + O2 + U3 + O3+ U4 + O4+ U5 + O5
ST
X11 + X12 + X13 + X14 + X15 + X16 + X17 + X18 + X19 + X1,10 + U1 – O1 =210
X21 + X22 + X23 + X24 + X25 + X26 + X27 + X28 + X29 + X2,10 + U2 – O2 =210
X31 + X32 + X33 + X34 + X35 + X36 + X37 + X38 + X39 + X3,10 + U3 – O3 =210
X41 + X32 + X43 + X44 + X45 + X46 + X47 + X48 + X49 + X4,10 + U4 – O4 =210
X51 + X52 + X53 + X54 + X55 + X56 + X57 + X58 + X59 + X5,10 + U5 – O5 =210
X11 + X21 + X31 + X41 + X51 = 1
X11 + X21 + X31 + X41 + X51 = 1
X12 + X22 + X32 + X42 + X52 = 1
X13 + X23 + X33 + X43 + X53 = 1
X14 + X24 + X34 + X44 + X54 = 1
X15 + X25 + X35 + X45 + X55 = 1
X16 + X26 + X36 + X46 + X56 = 1
X17 + X27 + X37 + X47 + X57 = 1
X18 + X28 + X38 + X48 + X58 = 1
X19 + X29 + X39 + X49 + X59 = 1
X1,10 + X2,10 + X3,10 + X4,10 + X5,10 = 1
Ui , Oi > 0
Xij binary
11.
Year
1 2 3 4 5
Minimum
Capacity in MW 780 860 950 1060 1180
ST X1 + X2 + X3 + X4 + X5 + X6 = 1800
X1 - 500 Y1 0
X2 - 600 Y2 0
X3 - 750 Y3 0
X4 - 400 Y4 0
X5 - 600 Y5 0
X6 - 800 Y6 0
Xi 0
All Yi are binary
MAX CijXij
S.T. Xij = 1, for each item i
CijXij 30, for each year j
17. a. Let Xij = bushels (in 1000s) shipped from grove i to processing plant j
Yij = 1 if Xij 0, 0 otherwise
ST X0 + X1 + X2 + X3 40
500X0 + 700X1 + 800X2 + 1,000X3 40,000
0 X0 40
5 X1 15
8 X2 22
0 X3 10
Xi integer
MIN X1 + X2 + X3 + X4 + X5 + X6
ST 3X1 + 2X2 + 2X3 + 1X4 + 0X5 + 0X6 5,000
0X1 + 1X2 + 0X3 + 2X4 + 1X5 + 0X6 1,200
0X1 + 0X2 + 1X3 + 0X4 + 1X5 + 2X6 300
Xi 0 & integer
b. See file: Prb6_19.xlsm
c. X1 = 1267, X3 = 300, X4 = 600, Total boards cut = 2167 (other solutions exist)
21. a. Let Pij = production at plant i allocated to meet demand for customer j
Yi= 1 if any Xij 0, 0 otherwise
Yi binary
22. a. Additional Constraints: Y1+Y2 1 and Y4+Y5 1 (implement as Y1 1-Y2 and Y4 1-Y5)
b. See file: Prb6_22.xlsm
c. The solution is: Build plants at locations 1, 3 and 4 (i.e., Y 1=Y3=Y4 = 1).
P1X=15,000, P1Y=25,000, P3X=5,000, P3Z=35,000, P4X=20,000.
Total cost = $7,800,000.
24. a. Let X1j = barrels of crude to buy from TX allocated to the production of product j
X2j = barrels of crude to buy from OK allocated to the production of product j
X3j = barrels of crude to buy from PA allocated to the production of product j
X4j = barrels of crude to buy from AL allocated to the production of product j
Yi = 1 if Xi > 0 and 0 otherwise
25. a. MIN S1 + S2 + S3
S.T. X11 + X12 + X13 + X14 + X15 + S1 = 2700
X21 + X22 + X23 + X24 + X25 + S2 = 3500
X31 + X32 + X33 + X34 + X35 + S3 = 4200
Y11 + Y21 + Y31 1
Y12 + Y22 + Y32 1
Y13 + Y23 + Y33 1
Y14 + Y24 + Y34 1
Y15 + Y25 + Y35 1
Xi1 - 2500Yi1 0, i=1 to 3
Xi2 - 2000Yi2 0, i=1 to 3
Xi3 - 1500Yi3 0, i=1 to 3
Xi4 - 1800Yi4 0, i=1 to 3
Xi5 - 2300Yi5 0, i=1 to 3
Yj binary
Xi 0
A - M YA 0
B - M YB 0
C - M YC 0
D - M YD 0
E - M YE 0
A - 50000 YA 0
B - 50000 YB 0
C - 50000 YC 0
D - 50000 YD 0
E - 50000 YE 0
A, B, C, D, E, M1, M2, M3 0
Yi binary
MIN 125*(N1+ N2+ N3+ N4) + 50*(Y1+ Y2+ Y3+ Y4) + 15*(E1+ E2+ E3+ E4)/100
ST Bi=Ei-1, i=2,3,4
Ei = Bi + 100*Ni - Di , i=1,2,3,4
Ni - M Y i 0
Ei 0
Yi binary
Ni 0 and integer
MAX 21 Y1 + 35 Y2 + 15 Y3 + 60 Y4 + 20 Y5 + 37 Y6
ST X1 + X2 + X3 + X4 + X5 + X6 =2
Y1 < 1X1 + 1X2 + 1X3 + 0X4 + 0X5 + 0X6
Y2 < 1X1 + 1X2 + 0X3 + 0X4 + 0X5 + 0X6
Y3 < 1X1 + 0X2 + 1X3 + 1X4 + 1X5 + 0X6
Y4 < 0X1 + 0X2 + 1X3 + 1X4 + 0X5 + 0X6
Y5 < 0X1 + 0X2 + 1X3 + 0X4 + 1X5 + 1X6
Y6 < 0X1 + 0X2 + 0X3 + 0X4 + 1X5 + 1X6
Xi binary, Yi binary
c.
37. a. See file Prb6_37.xlsm. Place hubs in Andersen, Berekley, Dillon, Edgefield, Hampton,. Kershaw,
Orangeburg & Union. Total hubs = 8, Total coverage = 55.
b. See file Prb6_37.xlsm. Place hubs in Bamberg, Colleton, Florence, Greenville, Greenwood, Hampton,
Kershaw, Marion, Orangeburg & Union. Total hubs = 10, Total coverage = 77.
38.
X 1= 2.25
X 2= 1.50
X 1<= 2 Obj = 25.5 X 1>= 3
X 1= 2.0 X 1= 3.0
X 2= 1.67 X 2= 0.0
Obj = 25.3 Obj = 18
X 2<= 1 X 2>= 2
X 1= 2.0 X 1= 1.5
X 2= 1.0 X 2= 2.0
Obj = 20 Obj = 25.0
X 1<= 1 X 1>= 2
X 1= 1.0
X 2= 2.33
Infeasible
Obj = 24.6
X 2<= 2 X 2>= 3
X 1= 1.0 X 1= 0.0
X 2= 2.0 X 2= 3.0
Obj = 22.0 Obj = 24.0
optimal solution
Chapter 6 - Integer Linear Programming : S-14
————————————————————————————————————————————
39. Candidate problems could be selected on a depth first, or breadth first basis. Other strategies might include
selecting the candidate with the best relaxed objective function value, or the lowest sum of integer
infeasibilities. All of these strategies are heuristic.
MIN 5 (X11 + X12 + … + X17) + 4 (X21 + X22 + … + X27) + 7 (X31 + X32 + … + X37) +
200 (R11 + R12 + … + R17) + 300 (R21 + R22 + … + R27) + 250 (R31 + R32 + … + R37) +
800 (S11 + S12 + … + S17) + 1000 (S21 + S22 + … + S27) + 700 (S31 + S32 + … + S37)
4. Old Dominion's marginal cost of producing the last 300 MW of the 4300 MW demanded on day 1 is
$3,050. (This result is obtained by reducing the demand on day 1to 4000 MW, re-optimizing the problem,
and subtracting the objective value ($139,700) from the value obtained in question 2.) If they can buy 300
MW for less than $3050, they should do so and not operate the James River plant on day 1.
5. This plan has the Galax and New River plants running at or near capacity each day. If demand is a little
higher than projected, it would be wise to keep the James River plant online continuously as well to avoid
several start ups.
Chapter 6 - Integer Linear Programming : S-15
————————————————————————————————————————————
3. $54,700
4. 250 = lots!
5. Increase capacity at site 2. Cost savings = $54,700-$48,968 = $5,732