An Improved Artificial Fish Swarm Algorithm Based on Chaotic Search and Feedback Strategy

Abstract

Artificial fish swarm algorithm (AFSA) is a kind of swarm intelligence algorithms, which has the features of not strict to parameter setting, insensitive to initial values, strong robustness and so on. But the precision can not be very high and artificial fish (AF) often suffers the problem of being trapped in local optima. Especially when the objective function is a multimodel function, this problem is more prominent. Since chaotic mapping enjoys certainty, ergodicity and stochastic property, chaotic search can serve as a kind of method for global optimization. Feedback can also act as a strategy to lead the movement of AF. In this paper, chaotic search and feedback strategy are introduced into AFSA to overcome the shortcoming above. The experimental results show that the improved AFSA can obtain better results than the standard AFSA.