A Modified Particle Swarm Optimization Algorithm

Aiming at the weaknesses of Particle Swarm Optimization algorithm (PSO) such as the easiness to get trapped into local optimal, slow convergence speed, and low precision, an improved PSO algorithm called Selectivity-Particle Swarm Optimization algorithm (S-PSO) is proposed in this paper. Differently from traditional PSO, according to S-PSO particles can refer not only to the optimal strategy, but can also select a better strategy based on their own consideration and then imitate it. When selecting local better strategy, S-PSO is introduced to increase particles’ intelligent behaviors. Based on the authority of strategy, the overlay group of strategy and spatial distance between particles themselves and the overlay group, particles will select the local better strategy. Experimental simulation results show that the global searching ability and convergence speed and precision of S-PSO algorithm is much better than traditional PSO.