A Multiobjective Brain Storm Optimization Algorithm Based on Decomposition

Pseudocode 2

The pseudo code of the algorithm MBSO/D.

Input:

MOP (1)

A stopping criterion

: the number of direction vectors (the clusters)

: the size of the cluster

: the number of the neighborhood,

: a set of uniformly distributed weight vectors

Output: Objective vectors:

Initialization: Generate an initial population determine ; randomly cluster

the initial population into clusters with size and determine the

best solution of each cluster; determine , where

are the closest weight vectors to ; set

While the function evaluation times are less than the maximum function evaluation times do

Set and

For , do

According to Eq. (4), randomly select two indexes and , use , and to generate offspring

by Eq. (5).

If dominates , is regenerated by Eq. (6).

Update of : For , if , then set

Update of Neighboring Solutions: set and

While and

;

If

randomly select a number from , set ,

, and .

.

If generated by one cluster

;

Else

;

End if

End if

;

End while

End for

Use Eq. (3) to update .

end for

end while