Example: Fighter Aircraft Problem – Linear
Assignment Method
From the original decision matrix
Fighter Aircraft Problem we can obtain attributewise preferences:
Rank |
f1 |
f2 |
f3 |
f4 |
f5 |
f6 |
weights |
0.2 |
0.1 |
0.1 |
0.1 |
0.2 |
0.3 |
1st |
A2 |
A2 |
A3 |
A3 |
A3 |
A1 |
2nd |
A4 |
A3 |
A1 A4 |
A5 |
A1 A4 |
A3 |
3rd |
A1 |
A4 |
|
A1 |
|
A2 A4 |
4th |
A3 |
A1 |
A2 |
A2 |
A2 |
|
Three criteria f3,f5,
and f6 have tied attributewise rankings. These can be equalised:
Rank |
f31f32 |
f51f52 |
f61f62 |
1st |
A3 A3 |
A3 A3 |
A1 A1 |
2nd |
A1 A4 |
A1 A4 |
A3 A3 |
3rd |
A4 A1 |
A4 A1 |
A2 A4 |
4th |
A2 A2 |
A2 A2 |
A4 A2 |
Each of these rankings gets half of
weight of the tied ranking and the matrix is:
|
1st |
2nd |
3rd |
4th |
A1 |
0.3 |
0.15 |
0.45 |
0.1 |
A2 |
0.3 |
0 |
0.15 |
0.55 |
A3 |
0.4 |
0.4 |
0 |
0.2 |
A4 |
0 |
0.45 |
0.40 |
0.15 |
The LP formulation is:
The optimal solution is a permutation
matrix P*:
The optimal order is than: (A3,A4,A1,A2).