SCD HW1 Solution Rev2

You might also like

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 12

Question 1:

Sunchem, a manufacturer of printing inks, has five manufacturing plants worldwide. Their
locations and capacities are shown in Table 5-7 along with the cost of producing 1 ton of ink at
each facility. The production costs are in the local currency of the country where the plant is
located. The major markets for the inks are North America, South America, Europe, Japan, and
the rest of Asia. Demand at each market and transportation costs from each plant to each market
in U.S. dollars are shown in Table 5-7. The expected exchange rates are shown in Table 5-8.
Management must come up with a production plan for the next year.
a. Build a Mixed-Integer Linear Programming model which minimize the total cost based
on this problem (Indices, Parameters, Decision Variables, Objective function, Constraints
should be also included).
b. How much should each plant produce and which markets should each plant supply? What
is the total cost for this production strategy?
c. If no plant can run below 50 percent of capacity, which constraints should be added or
revised? Then, how much should each plant produce and which markets should each
plant supply? What is the total cost for this production strategy?

a. MILP model:
● Indices:
i: index of Market (i=1,…,5)
j: index of Plant (j=1,…,5)
● Parameters:
Tij: Transportation cost from Plant i to Market j (USD)
Ci: Capacity of Plant i (Tons)
Pi: Production cost per Ton at Plant i (local currency)
Ei: Exchange rate of local currency at Plant i to USD
Dj: Demand at Market j (Tons)
● Decision variables:
xij: Production quantity of demand at Market j assign to Plant i
yij: Binary, =0 if some demand at Market j assign to Plant I, =0 if otherwise
● Objective function:
i j i j
Minimize ∑ ∑ T ij × xij + ∑ ∑ Pi × Ei × x ij
1 1 1 1
or
i j i j
Minimize ∑ ∑ T ij × y ij + ∑ ∑ P i × E i × xij
1 1 1 1

 Constraints:
i
(1) ∑ x ij=D j , for j=1 , … , 5
1
j
(2) ∑ x ij ≤ K i , for i=1 , … , 5
1

● CPLEX Code:

//Indices:
int numPlant=5;
int numMarket=5;
range Plant=1..numPlant;
range Market=1..numMarket;

//Parameters:
float T[Plant][Market]=[[600,1300,2000,1200,1700],
[1300,600,1400,1400,1300],
[2000,1400,300,2100,900],
[1200,1400,2100,800,2100],
[2200,1300,1000,2300,800]];
float C[Plant]=[185,475,50,200,80];
float P[Plant]=[10000,15000,1800000,13000,400000];
float E[Plant]=[1,0.502,0.0093,0.562,0.023];
float D[Market]=[270,200,120,190,100];

//Decision Variables:
dvar int+ x[Plant][Market];

//Objective function:
minimize sum(i in Plant,j in Market)T[i][j]*x[i][j] + sum(i in Plant,j in Market)P[i]*E[i]*x[i][j];
//Constraints:
subject to{
//Constraint 1:
forall(j in Market){
sum(i in Plant)x[i][j] == D[j];
}

//Constraint 2:
forall(i in Plant){
sum(j in Market)x[i][j] <= C[i];
}
}

b. The optimal order strategy:

x[i][j] Market, j
1. North 4. South
Plant, i America 2. Europe 3. Japan America 5. Asia
1. United
States 125 0 0 0 0
2. Germany 135 200 120 0 20
3. Japan 0 0 0 0 0
4. Brazil 10 0 0 190 0
5. India 0 0 0 0 80

● Total cost: USD 7,816,450


● CPLEX code:
● //Indices:
● int numPlant=5;
● int numMarket=5;
● range Plant=1..numPlant;
● range Market=1..numMarket;

● //Parameters:
● float T[Plant][Market]=[[600,1300,2000,1200,1700],
● [1300,600,1400,1400,1300],
● [2000,1400,300,2100,900],
● [1200,1400,2100,800,2100],
● [2200,1300,1000,2300,800]];
● float C[Plant]=[185,475,50,200,80];
● float P[Plant]=[10000,15000,1800000,13000,400000];
● float E[Plant]=[1,0.502,0.0093,0.562,0.023];
● float D[Market]=[270,200,120,190,100];

● //Decision Variables:
● dvar int+ x[Plant][Market];

● //Objective function:
● minimize sum(i in Plant,j in Market)T[i][j]*x[i][j] + sum(i in Plant,j in Market)P[i]*E[i]*x[i][j];

● //Constraints:
● subject to{
● //Constraint 1:
● forall(j in Market){
● sum(i in Plant)x[i][j] == D[j];
● }

● //Constraint 2:
● forall(i in Plant){
● sum(j in Market)x[i][j] <= C[i];
● sum(j in Market)x[i][j] >= (1/2)*C[i];
● }
● }
c. Constraint (2) should be revised as below:
j
1
× K ≤ ∑ xij ≤ K i , for i=1, … , 5
2 i 1

The optimal order strategy:


x[i][j] Market, j
1. North 4. South
Plant, i America 2. Europe 3. Japan America 5. Asia
Plant, i 1 2 3 4 5
1. United
States 100 0 0 0 0
2. Germany 160 200 95 0 20
3. Japan 0 0 25 0 0
4. Brazil 10 0 0 190 0
5. India 0 0 0 0 80

● Total cost: USD 7,974,950

Question 2: (Dataset Revised)

Given the dataset of a supply-demand network in the following table:

Location X co-ordinates Y co-ordinates Supply or demand


Supplier 1 61 44 40
Supplier 2 12 6 16
Supplier 3 38 26 76
Supplier 4 6 22 72
Warehouse 1 63 22 54
Warehouse 2 26 8 77
Warehouse 3 18 62 37
Warehouse 4 8 65 43
Warehouse 5 43 36 26
Warehouse 6 8 41 68

a. Use Center of Gravity (COG) method to find a new location for building new
Distribution Center without cost in consideration.
b. Given the transportation costs of Supplier i={1,2,3,4} and Warehouse j={1,2,3,4,5,6} are
respectively ($/Ton Mile): (1.4,0.4,1.6,0.3,1.4,0.7,0.6,0.3,1.8,1.7). Calculate the Total
Transportation Cost of this supply-demand network.
c. Use Excel Solver to find a new location for building new Distribution Center to minimize
the Total Transportation Cost.

Solution:

a.

∑ XiW i ∑Y i W i
X 0= ; Y 0=
∑W i ∑W i

Location X Y co-ordinates Supply or demand X*Weigh Y*Weigh


coordinates (Weight) t t
Supplier 1 61 44 40 2440 1760
Supplier 2 12 6 16 192 96
Supplier 3 38 26 76 2888 1976
Supplier 4 6 22 72 432 1584
Warehouse 1 63 22 54 3402 1188
Warehouse 2 26 8 77 2002 616
Warehouse 3 18 62 37 666 2294
Warehouse 4 8 65 43 344 2795
Warehouse 5 43 36 26 1118 936
Warehouse 6 8 41 68 544 2788
Total 283 332 509 14028 16033

∑ X i W i 14028
X0 = = = 27.56
∑Wi 509
∑ Y i W i 16033
Y0 = = = 31.5
∑W i 509
A good place to start looking for locations is around (27.56, 31.5). No warehouse is close to this
location, so it might be better to look for an entirely new location at (X0, Y0) =(27.56, 31.5).

b.
To calculate the distance, we apply Euclidean Distance:
2

Distance= ( x −xn ) + ( y− y n )
2

Location X co- Y co- Supply or demand $/Ton Distance Cost


ordinates ordinates (W) Mile
Supplier 1 61 44 40 1.4 35.700327 1999.218
9
Supplier 2 12 6 16 0.4 29.871576 191.1781
1
Supplier 3 38 26 76 1.6 11.799764 1434.851
2
Supplier 4 6 22 72 0.3 23.559744 508.8905
2
Warehouse 1 63 22 54 1.4 36.691014 2773.841
Warehouse 2 26 8 77 0.7 23.550736 1269.385
4
Warehouse 3 18 62 37 0.6 31.964073 709.6024
9
Warehouse 4 8 65 43 0.3 38.793122 500.4313
4
Warehouse 5 43 36 26 1.8 16.082750 752.6727
7
Warehouse 6 8 41 68 1.7 21.745325 2513.76
7
  12653.83
0

c.
Location X co- Y co- Supply or $/Ton Mile Distance Cost
ordinates ordinates demand
Supplier 1 61 44 40 1.4 29.20617695 1635.54
6
Supplier 2 12 6 16 0.4 32.80242568 209.935
5
Supplier 3 38 26 76 1.6 1.38865E-05 0.00168
9
Supplier 4 6 22 72 0.3 32.24901747 696.578
8
Warehous 63 22 54 1.4 25.31798991 1914.04
e1
Warehous 26 8 77 0.7 21.63329643 1166.03
e2 5
Warehous 18 62 37 0.6 41.18251843 914.251
e3 9
Warehous 8 65 43 0.3 49.20365426 634.727
e4 1
Warehous 43 36 26 1.8 11.18035 523.240
e5 4
Warehous 8 41 68 1.7 33.54101015 3877.34
e6 1
Total 11571.7

Optimal x 38.00
Optimal y 26.00

Question 3: (Dataset Revised)

Given the dataset of a supply network in the following table:

Location X co-ordinates Y co-ordinates Supply or demand


Site 1 61 44 40
Site 2 12 6 16
Site 3 38 26 76
Site 4 6 22 72
Supplier 1 63 22 54
Supplier 2 26 8 77
Supplier 3 18 62 37
Supplier 4 8 65 43
Supplier 5 43 36 26
Supplier 6 8 41 68

a. Calculate the Euclidean distance from every potential site to every supplier
b. Calculate the load-distance value for every potential site.
c. Conclude on the best site with the lowest load-distance value.
For example:
Distance from site 1 to supplier 1 = √ 63−61¿2 + ( 22−44 )2 ¿=22.09

Location Supplie Supplie Supplie Supplie Supplie Supplie


r1 r2 r3 r4 r5 r6
Site 1 22,09 50,21 46,62 57,01 19,70 53,08
Site 2 53,45 14,14 56,32 59,14 43,14 35,23
Site 3 25,32 21,63 41,18 49,20 11,18 33,54
Site 4 57,00 24,41 41,76 43,05 39,56 19,10

b.

For site 1:
LD = 22.09(54) + 50.21(77) + 46.62(37) + 57.01(43) + 19.70(26) + 53.0868) = 13357.09

n
LD = ∑ I i d i
i=1

Site 1 Site 2 Site 3


Site 4
Supplier 1 22.091 53.451 25.318
57.000
Supplier 2 50.210 14.142 21.633
24.413
Supplier 3 46.615 56.321 41.183
41.761
Supplier 4 57.009 59.135 49.204
43.046
Supplier 5 19.698 43.139 11.180
39.560
Supplier 6 53.085 35.228 33.541
19.105
Load Distance 13357.093 12119.091 9243.924
10681.674
c.
Site 3 has the lowest load-distance value, therefore choose site 3 as the best site.

Question 4:

Beika Tower Shopping Mall would like to build up a new Distribution Center to deliver orders to
6 cities, with locations, distances and demands shown in the following figure.

a. Write off the distance matrix between these locations.


b. Find the Single Median with the lowest total cost of weight-distance.

City Demand

A 90

B 120

C 110

D 65

E 80

F 95

a.

To A B C D E F Demand
From
A 0 3 2 8 10 13 90
B 3 0 1 5 7 10 120
C 2 1 0 6 8 11 110
D 8 5 6 0 2 5 65
E 10 7 8 2 0 3 80
F 13 10 11 5 3 0 95

b.
For example : Weight – distance of city A is: 90*0 + 120*3 + 110*2 + 65*8 + 80*10
+ 95*13 = 3135

B has the lowest total cost, therefore we choose city B

Question 5: (Hint added)

Beika Tower Shopping Mall would like to build up a new Distribution Center to deliver orders to
6 cities, with locations, distances and demands shown in the following figure.

a. Calculate the distance matrix.


b. Where should you locate a logistics center that gives the shortest travel distance over the
network?
c. Where should they build the minimum number of depots to give maximum travel
distance of 3 kilometers to any other city?

(Hint: Consider this question as a Set Covering Problem)


City Demand

A 90

B 120

C 110

D 65

E 80

F 95

a.

b In conclusion, logistics center should locate at city D to give the shortest travel distances over
the network.
c. To give maximum travel distance of 3 kilometers to any other city, the possible distance
matrix is:

  A F C B E
A 0   2 3  
B 3   1 0  
C 2   0 1  
D         2
E   3     0
F   0     3
Maximu 3 3 2 1 3
m

 We can choose E and A, E and B, E and C with the maximum travel time of 3
kilometers

You might also like