Stochastic Particle Swarm Optimization Algorithm

Particle swarm optimization(PSO) is a new heuristic global optimization algorithm based on swarm intelligence after ant colony algorithm.The algorithm is simple,easy to implement and has good performance of optimization.Now it has been applied in many fields.However,when optimizing multidimensional and multimodal functions,the basic particle swarm optimization is apt to be trapped in local optima.This paper proposes a modified optimization method——stochastic particle swarm optimization(SPSO),which combines the standard version with simulated annealing algorithm.This modified version has stronger exploitation ability at the beginning,so it can keep particle swarm from getting into local optima too early.Simulation results on benchmark complex functions with high dimension show that this hybrid algorithm performs better than the basic particle swarm optimization.