关键词: 人工鱼群算法, 自适应, 动态邻域, 粒子群算法

Abstract: Aiming at the problems of easily falling into local optimum and low optimization precision in the Artificial Fish Swarm Algorithm（AFSA）, an Artificial Fish Swarm Algorithm based on Adaptive Dynamic Neighborhood Structure（ADAFSA） is proposed. In the algorithm, the dynamic neighbors of each artificial fish are adaptively constructed according to the distance between the fish with the others and the current iteration, then the visual and the step of each artificial can be adaptively calculated with the dynamic neighborhood. The algorithm also adds information strategy of Particle Swarm Optimization（PSO） and billboard to improve behaviors of artificial fish. The simulation results show that the ability of the proposed algorithm overcomes the local optimum to achieve stronger global optimization, and optimization precision is higher.

Key words: Artificial Fish Swarm Algorithm（AFSA）, adaptively, dynamic neighborhood, Particle Swarm Optimization（PSO）