The Parellel Particle Swarm Optimization Based on Dynamic Distance Threshold

Through introducing dynamic distance threshold to standard PSO, the particles are divided into three categories of the vicinity of the best position, the vicinity of the average position and the other position, which are formed three Subspecies. The first category execute concentrated search, the second category do general search and the third category do scatter search, which reasonably balance the contradiction of coarse search and detailed search. This bring about that convergence speed is improved under the circumstance of keeping fundamental stable condition in particle diversity. The parallel algorithm is used to speed up calculating speed, to keep the particle diversity and to jump out local optimization. The simulating experiments have certified that the algorithm can improve not only the convergence but also the particle diversity.