摘要
保持遗传算法在演化过程中的种群多样性 ,是将遗传算法成功应用于解决多峰优化问题和多目标优化问题的关键。适应值共享遗传算法和拥挤遗传算法分别从不同角度改善了遗传算法的搜索能力 ,是寻找多个最优解的常用算法。将这两种算法的优点加以结合 ,提出适应值共享拥挤遗传算法。数值测试结果表明 。
By combining fitness sharing method and crowding method in selection stage and replacement stage of genetic algorithms respectively, a fitness sharing crowding genetic algorithms is proposed. It combines the advantages of both fitness sharing genetic algorithms and crowding genetic algorithms in searching ability. The suggested algorithm is proved to be more efficient in solving benchmark multimodal optimization problems than standard fitness sharing genetic algorithm and crowding genetic algorithm.
出处
《控制与决策》
EI
CSCD
北大核心
2001年第6期926-929,共4页
Control and Decision
关键词
遗传算法
适应值共享
多峰优化
genetic algorithms
fitness sharing
crowding
multimodal optimization
