DISTRIBUTION MODELS

The Transportation Model

deals with the shipment of a commodity from a number of sources to a number of destinations sources are referred to as the supply points, while the destinations are also called the demand points the availability at each supply point is fixed limited and there is a specific order quantity at each demand point the objective is to determine the amount shipped from each supply point to each demand point so that total transportation cost is minimized

Structure of a Transportation Problem

All constraints are equality constraints. The total availability at the supply points is equal to the total requirement of the demand points. Each coefficient in the constraints matrix is either 1, -1, or 0. Each column of the constraints matrix, except for the column on the right side of the equality sign, has exactly two nonzero elements.

Sample Problem: Sony Co., a manufacturer of TV sets, has three manufacturing plants and three main distributors. The per-unit costs of shipping from plant 1 to distributors 1, 2, and 3 are P60, P80, and P50, respectively. The corresponding costs of shipping from plant 2 are P90, P20, and P40. The costs from plant 3 are P40, P70, and P60. Next month, the company expects to have 250 TV sets available at plant 1, 225 sets available at plant 2, and 425 sets available at plant 3. The monthly requirements of distributors 1, 2, and 3 are 200, 300, and 400 sets, respectively. The objective is to find the least-cost way of shipping from the three plants to the three distributors. Supply points P1 P2 P3 Demand points D1 D2 D3

Mathematical Model
x11 = the number of TV sets to be shipped from plant 1 to distributor 1

x12= the number of TV sets to be shipped from plant 1 to distributor 2 x13= the number of TV sets to be shipped from plant 1 to distributor 3 x21= the number of TV sets to be shipped from plant 2 to distributor 1 x22= the number of TV sets to be shipped from plant 2 to distributor 2 x23= the number of TV sets to be shipped from plant 2 to distributor 3 x31= the number of TV sets to be shipped from plant 3 to distributor 1 x32= the number of TV sets to be shipped from plant 3 to distributor 2 x33= the number of TV sets to be shipped from plant 3 to distributor 3
Minimize C = 60x11+80x12+50x13+90x21+20x22+40x23+40x31+70x32+60x33 Subject to x11+x12+x13 = 250 x21+x22+x23 = 225 +x31+x32+x33 = 425 x11 +x21 +x31 = 200 x12 +x22 +x32 = 300 x13 +x23 +x33 = 400

Tabular representation of the Problem

Distributor

1
60

2
80

3
50

Availability

1
Plant

x11
90

x12
20

x13
40

250 225
60

2 3
Requirement

x21
40

x22
70

x23 x33 400

x31 200

x32 300

425

The Transportation Method

A search and evaluation algorithm similar to the solution of ordinary linear progg problems General steps for solving minimization transportation model are as follows:
1. 2. 3. 4. Present the problem in tabular form Find an initial feasible solution (using Northwest corner method) Calculate the net contributions of all the nonbasic variables in the current solution (using the stepping stone method or MODI method). Test the current solution for optimality. To do this, simply examine the algebraic signs of the net contributions obtained in step2. If no negative signs exists, then the current solution is optimal. If at least one negative sign exist, then move on to the next step. Generate an improved solution Go back to step3.

5. 6.

The Northwest Corner Method - a method of finding an initial solution to a transportation model. Steps 1. Begin in the upper left-hand cell (northwest corner) of the main body of the transportation table. Allocate the smaller amount of either the row supply or the column demand. 2. If the supply constraint is satisfied, go to step 3; otherwise move on to the next cell on the right. Allocate into this cell the largest value that exceeds neither the remaining supply for that row nor the demand for that column. If supply is still not exhausted, repeat this step until the supply constraint is satisfied. 3. When the supply for a particular row has been satisfied, move down to the next cell in the same column. 4. Repeat steps 2 and 3 until all the supply and demand constraints are satisfied.

 In any transportation problem, number of used cells = no. of rows + no. of columns 1 Stepping Stone Method - to evaluate the net contributions of the unused cells. - to determine what would happen to the total objective function value if one unit (of product) was shipped over a route that is presently not being used.

Steps in Stepping Stone Method:

1. Beginning with the unused cell to be evaluated, trace a path that forms a closed loop using only used cells as corners. The closed loop may skip over either used or unused cells and may, at times, cross over itself. The number of cells in any particular closed loop must be an even number of at least four. 2. Assign a + sign to the unused cell being evaluated and then alternately assign a sign and a + sign on each of the remaining corners of the loop. 3. Calculate the net contribution of the unused cell by adding the per unit-costs of all cells containing a + sign and subtracting the per-unit costs of all cells containing a sign. The net contribution value may be either positive, negative, or zero, indicating an increase, decrease, or no change in the total cost. For minimization problem, the solution is optimal if the net contributions are either zero or positive.

Steps in Stepping Stone Method:

Developing an improved solution: 1. Select a route to be utilized. In minimization, select the route or cell with the smallest (most negative) net contribution value. If a tie exists for the smallest value, select either. 2. Determine the largest amount to be shipped over the selected route. The maximum amount that can be allocated to the selected route is equal to the smallest quantity with a sign in the closed loop. 3. Add the amount determined in step 2 to the quantities with a + sign and subtract it from the quantities with a sign in the closed. Quantities not affected by the closed loop are carried forward as is to the next table.

 The MODI Method

The modified distribution method of MODI method is an alternative method of evaluating the net contributions of the unused cells in the transportation table. It is computationally more efficient than the stepping stone method. Steps: 1. Denote the rows by R1, R2, R3, . . . and the columns by K1, K2, K3, ... 2. Set R1 = 0 and compute the values of all Kjs and the remaining Ris by applying the formula Ri + Kj = cij on the used cells, where Ri = the value of the ith row Kj = value of the jth column cij = per-unit cost of the used cell in the ith row, jth column 3. Calculate the net contributions of the unused cells by the formula NC = cij (Ri + Kj), where NC = net contributions of the unused cells cij = per-unit cost of the unused cell in the ith row, jth column

Table 1 K1 60 K2 80 K3 50 Availability

R1 R2

200
90

50
20 40

250 225
70 60

225
40

R3
Requirement

25 200 300

400 400

425

There are 5 used cells and so there are also equations of the form Ri + Kj=cij R1 + K1 = 60 R1 + K2= 80 R2 + K2= 20 R3 + K2= 70 R3 + K3= 60

If R1= 0, we obtain the ff solutions R1 = 0 R2 = -60 R3 = -10 K1 = 60 K2 = 80 K3 = 70

 Unbalanced Transportation Problem

A transportation problem is said to be balanced if the total availability at the supply points is equal to the total requirement of the demand points. A transportation problem is unbalanced if supply is greater than demand or demand is greater than supply. If supply is greater than demand, add one extra destination (or a dummy demand point) to balance the problem. An additional column is placed in the transportation table. The excess supply is then entered as its requirement. Allocation to its cells are simply not shipped anywhere, hence, it makes sense to assign a per-unit cost of zero to the cells in the additional column. If demand is greater than supply, add extra origin or a dummy supply point. The excess demand then becomes the availability for the additional row in the table. Allocation to cells associated with the dummy supply point never arrive to any destination, hence, a per-unit cost of zero to these additional cells.

 Degeneracy in the Transportation Method

If the number of used cells in any solution is less than the sum of rows and columns less one, then degeneracy exists and the net contributions of the unused routes cannot be evaluated by standard methods. To resolve degeneracy, assign a usable zero to one of the previously unused cells. Assignment of this usable zero converts an unused cell to a used cell and allows us to include that cell in the computations of the net contributions. More than one usable zero is necessary to obtain the required number of used cells.

Initial Solution to an Unbalanced Problem Demand is less than Supply Distributor

A
5

B
4 4

C
3

Availability

D
Plant

250
8 3

250 200
9 7

E F
Requirement

50

50
5

300 300

150 300 200 200

Initial Solution to an Unbalanced Problem Demand is less than Supply Distributor

Dummy Availability warehouse

A
5

B
4 4

C
3 3

D
Plant

250
8 0

250 200
9 7

E F
Requirement

50

50
5 0

300 150 150 300

150 300 200 200

Initial Solution to an Unbalanced Problem Demand is greater than Supply Warehouse Plant Supply 3

A
5

B
4 4

W
Plant

200
8 3

200 100
9 7

X Y
Requirement

50

25
5

175 75

75 250 100 150

Initial Solution to an Unbalanced Problem Demand is greater than Supply Warehouse Plant Supply 3

A W
Plant 5

B
4 4

200
8 3

200 100
9 0 7

X Y Dummy Plant
Requirement

50

25
5

175 75
0

75
0

50 250 100 150

50