Transportation Optimality
1. **State the problem:** We are given a transportation problem with factories F1, F2, F3 supplying units to warehouses W1, W2, W3, W4. The costs per unit and supplies/demands are provided. A feasible solution with allocations is given, and we must:
(i) Test optimality of the solution using the MODI method.
(ii) If not optimal, find the optimal solution.
(iii) Determine the minimum transportation cost.
2. **Data summary:**
Cost matrix (C) per unit:
$$
\begin{array}{c|cccc}
& W1 & W2 & W3 & W4 \\
\hline
F1 & 12 & 6 & 20 & 25 \\
F2 & 6 & 11 & 15 & 12 \\
F3 & 9 & 15 & 17 & 7
\end{array}
$$
Supply vector:
$$[F1:800, F2:600, F3:1000]$$
Demand vector:
$$[W1:400, W2:500, W3:700, W4:800]$$
Allocations matrix (X):
$$
\begin{array}{c|cccc}
& W1 & W2 & W3 & W4 \\
\hline
F1 & 300 & 500 & 0 & 0 \\
F2 & 100 & 0 & 500 & 0 \\
F3 & 0 & 0 & 200 & 800
\end{array}
$$
3. **Step 1: Compute the initial transportation cost:**
Total cost $Z = \sum C_{ij} X_{ij}$
From allocations:
- F1 to W1: $12 \times 300 = 3600$
- F1 to W2: $6 \times 500 = 3000$
- F2 to W1: $6 \times 100 = 600$
- F2 to W3: $15 \times 500 = 7500$
- F3 to W3: $17 \times 200 = 3400$
- F3 to W4: $7 \times 800 = 5600$
Sum:
$$Z = 3600 + 3000 + 600 + 7500 + 3400 + 5600 = 23700$$
4. **Step 2: Apply MODI Method to test optimality:**
We define potentials $u_i$ for rows (factories) and $v_j$ for columns (warehouses) such that for every allocated cell $(i,j)$:
$$ u_i + v_j = C_{ij} $$
Set $u_1 = 0$ (arbitrary):
Allocated cells:
- $(F1, W1): u_1 + v_1 = 12 \Rightarrow 0 + v_1 = 12 \Rightarrow v_1=12$
- $(F1, W2): u_1 + v_2 = 6 \Rightarrow 0 + v_2 =6 \Rightarrow v_2=6$
- $(F2, W1): u_2 + v_1 = 6 \Rightarrow u_2 + 12 =6 \Rightarrow u_2 = -6$
- $(F2, W3): u_2 + v_3 = 15 \Rightarrow -6 + v_3=15 \Rightarrow v_3=21$
- $(F3, W3): u_3 + v_3=17 \Rightarrow u_3 + 21=17 \Rightarrow u_3=-4$
- $(F3, W4): u_3 + v_4=7 \Rightarrow -4 + v_4=7 \Rightarrow v_4=11$
5. **Step 3: Calculate opportunity cost (reduced cost) for non-allocated cells:**
For each non-allocated cell $(i,j)$,
$$ ext{Opportunity cost }=C_{ij} - (u_i + v_j) $$
- $(F1, W3): 20 - (0 + 21) = -1$
- $(F1, W4): 25 - (0 + 11) = 14$
- $(F2, W2): 11 - (-6 + 6) = 11$
- $(F2, W4): 12 - (-6 + 11) = 7$
- $(F3, W1): 9 - (-4 + 12) = 1$
- $(F3, W2): 15 - (-4 + 6) = 13$
6. **Step 4: Check optimality:**
If all opportunity costs $\\geq 0$, solution is optimal. Here, $(F1, W3)$ has negative opportunity cost $-1$ indicating solution is not optimal.
7. **Step 5: Improve solution by allocating units through closed loop adjustment:**
We choose cell $(F1, W3)$ to enter the basis.
Constructing a loop alternating + and - signs:
Path:
- Start: (F1, W3) [+]
- Down to (F2, W3) [-]
- Left to (F2, W1) [+]
- Up to (F1, W1) [-]
- Back to (F1, W3) to close loop
8. **Step 6: Find minimum allocation in cells with minus signs:**
Cells with minus signs: (F2, W3) = 500 units, (F1, W1) = 300 units
Minimum is $\theta = 300$
9. **Step 7: Update allocations along the loop:**
Add $\theta$ to cells with plus signs and subtract from cells with minus signs:
- (F1, W3): 0 + 300 = 300
- (F2, W3): 500 - 300 = 200
- (F2, W1): 100 + 300 = 400
- (F1, W1): 300 - 300 = 0
10. **Step 8: New allocations matrix:**
$$
\begin{array}{c|cccc}
& W1 & W2 & W3 & W4 \\
\hline
F1 & 0 & 500 & 300 & 0 \\
F2 & 400 & 0 & 200 & 0 \\
F3 & 0 & 0 & 200 & 800
\end{array}
$$
11. **Step 9: Calculate new total transportation cost:**
- F1 to W2: $6 \times 500 = 3000$
- F1 to W3: $20 \times 300 = 6000$
- F2 to W1: $6 \times 400 = 2400$
- F2 to W3: $15 \times 200 = 3000$
- F3 to W3: $17 \times 200 = 3400$
- F3 to W4: $7 \times 800 = 5600$
Sum:
$$Z = 3000 + 6000 + 2400 + 3000 + 3400 + 5600 = 23400$$
Cost reduced from 23700 to 23400.
12. **Step 10: Repeat MODI method to check if this new solution is optimal:**
Set $u_1=0$.
Allocate potentials:
- (F1, W2): $0 + v_2 = 6 \Rightarrow v_2=6$
- (F1, W3): $0 + v_3=20 \Rightarrow v_3=20$
- (F2, W1): $u_2 + v_1=6$
- (F2, W3): $u_2 + 20=15 \Rightarrow u_2= -5$
- So, $u_2 + v_1=6 \Rightarrow -5 + v_1=6 \Rightarrow v_1=11$
- (F3, W3): $u_3 + 20=17 \Rightarrow u_3 = -3$
- (F3, W4): $u_3 + v_4=7 \Rightarrow -3 + v_4=7 \Rightarrow v_4=10$
Check opportunity costs for non-allocated:
- (F1, W1): $12 - (0 + 11)=1$
- (F1, W4): $25 - (0 + 10)=15$
- (F2, W2): $11 - (-5 + 6)=10$
- (F2, W4): $12 - (-5 + 10)=7$
- (F3, W1): $9 - (-3 + 11)=1$
- (F3, W2): $15 - (-3 + 6)=12$
All non-allocated have $\\geq 0$, so solution is optimal.
**Final answer:**
Optimal allocations:
$$
\begin{array}{c|cccc}
& W1 & W2 & W3 & W4 \\
\hline
F1 & 0 & 500 & 300 & 0 \\
F2 & 400 & 0 & 200 & 0 \\
F3 & 0 & 0 & 200 & 800
\end{array}
$$
Minimum transportation cost: $$ \boxed{23400} $$