Page 57 - Algorithm Collections for Digital Signal Processing Applications using MATLAB
P. 57
1. Artificial Intelligence 45
Figure 1-25. Matrix A
Figure 1-26. Matrix B
The cost of the order [3 2 1 5 7 2 4 6 8] is computed as the sum of the
product of values obtained from the positions (1,3) or (3,1), (2,2), (3,1), (4,5)
or (5,4) , (5,7) or (7,5) ,(6,2), (7,4) and (8,6) of the matrices A and B .
Cost ([3 2 1 5 7 2 4 6 8] ) = a6*b6 + a3*b3 + a6*a6 + a14*b14 + a26*b26 +
a17*b17 +a25*b25 + a36*b36
(see Matrix A and Matrix B )
The problem of finding the optimal order is achieved using Ant colony
optimization technique as described below.
Step 1: Generate the 5 sets of orders randomly which are treated as the
initial values selected by the 5 Ants respectively.
The following are the orders selected by the 5 ants along with the
corresponding cost measured as described in the previous section.
