Improved Particle Swarm Optimization By Fast Simulated Annealing Algorithm

摘要

This paper proposes a hybrid particle swarm optimization with the fast-simulated annealing (PSO-FSA). The proposed algorithm is meant to solve high dimensional optimization problems based on two strategies, which are utilizing the particle swarm optimization to define the global search area and utilizing the fast-simulated annealing to refine the visited search area. To evaluate its performance, we examined the algorithm on 14 benchmark functions. Based on the results, PSO-FSA has higher accuracy result compared with particle swarm, simulated annealing. We also apply the algorithm in clustering problem, and the results shows that the proposed method has better accuracy than the optimization methods.