基于变异算子的人工鱼群混合算法

Hybrid artificial fish swarm algorithm based on mutation operator

把Hooke-Jeeves模式搜索方法作为人工鱼群算法的一个局部搜索算子,嵌入到带变异算子的人工鱼群算法中,提出一种基于变异算子的人工鱼群混合算法。其中,Hooke-Jeeves的强局部搜索能力提高了人工鱼群混合算法的局部收敛速度,变异算子的引入增加了群体的多样性,避免人工鱼群混合算法陷入局部最优。通过基准函数和实例测试验证,表明了该算法是高效可行的。

A Hybrid Artificial Fish Swarm Algorithm（HAFSA） is proposed,where the Hooke-Jeeves pattern search is combined with AFSA to speed up the local search,also mutation operator is embedded to avoid the common defect of premature convergence.So the hybrid algorithm is capable of improving the local search ability of the algorithm,as well as reducing the computational burden.The experimental results through testing by functions and practical problems show that the hybrid algorithm is efficient and effective.