Computational Intelligence

138 Chapter Four—Evolutionary Computation Implementations

{ // end of the cycle

cur_pso = 0; // move to the first pso

}

PSO_current_state = PSO_UPDATE_PBEST_EACH_CYCLE;

// move to the next state

psos[cur_pso].popu_index = 0;

}

}

The PSO UPDATE INERTIA WEIGHT State

In this state, the current PSO updates its inertia weight according to its inertia weight

updating method. The state transitions to the state

PSO_EVALUATE. The index of

the population is set to 0 to start from the first individual. The state handler routine

is shown in Listing 4.32.

Listing 4.32

The PSO UPDATE INERTIA WEIGHT state handler routine.

static void PSO_update_inertia_weight (void)

{

iw_update_methods(psos[cur_pso].env_data.iw_method); ...