Simulated annealing artificial fish swarm algorithm

Abstract

This paper presents a novel stochastic approach called the simulated annealing-artificial fish swarm algorithm (SA-AFSA) for solving some multimodal problems. The proposed algorithm incorporates the simulated annealing (SA) into artificial fish swarm algorithm (AFSA) to improve the performance of the AFSA. The hybrid algorithm has the following features: the hybrid algorithm maintains 1) the strong local searching ability of the SA and 2) the swarm intelligence of AFSA. The experimental results indicate that in all the test cases, the SA-AFSA can obtain much better optimization precision and the convergence speed compared with AFSA.