多峰函数优化的免疫人工鱼群网络算法

针对人工鱼群算法(artificial fish swarm algorithm, AFSA)多峰寻优能力不足的问题，提出了一种免疫人工鱼群网络算法。应用改进的觅食行为，提升了算法的局部寻优能力；采用免疫网络调节机理，保持了人工鱼群多样性，不〖JP2〗断探寻新的局部峰值；执行模式搜索法（pattern search method, PSM），完成精英人工鱼群的精细搜索。仿真实验结果表明，该算法具有较强全局优化能力和局部优化能力，且搜索到每个最优解都达到了理想值。

Immune artificial fish swarm network algorithm for multimodal function optimization

An immune artificial fish swarm network algorithm is proposed to deal with the problem of inefficient searching that the artificial fish swarm algorithm (AFSA) has difficulty in solving multimodal function optimization. In the algorithm, the local searching capacity can be enhanced by using improved preying behavior. Referred to immune network theory, the diversity of artificial fish swarm is maintained and new local solutions can be found continuously. The pattern search method (PSM) is introduced to obtain the local optimum solution by its good local extremum search ability. Simulation results show that the proposed algorithm has excellent global optimization performance and good local extremum search ability and can give satisfactory solutions.