An Improved Artificial Fish Swarm Algorithm

Abstract

The purpose of this paper is to improve the performance of the original AFSA algorithm at the optimal accuracy rate and overcome the weakness of the algorithm which is also trapped in the local optimum. To this end, the original AFSA was further improved based on the tabu strategy. Specifically, the reproduction and death were introduced to protect the best individuals and eliminate poor quality fish, so as to increase convergence and accuracy. Through simulation, it is proved that our solution can achieve high accuracy, good global convergence, and strong resistance to local minimum. The findings bring new light on the application of AFSA and provide valuable reference to studies in related fields.