一种基于阈值选择策略的改进混合蛙跳算法

混合蛙跳算法(SFLA)是一种全新的后启发式群体进化算法,具有高效的计算性能和优良的全局搜索能力。对混合蛙跳算法的基本原理进行了阐述,针对算法局部更新策略引起的更新操作前后个体空间位置变化较大,降低收敛速度这一问题,提出一种基于阈值选择策略的改进混合蛙跳算法。通过不满足阈值条件的个体分量不予更新的策略,减小了个体空间差异,从而改善了算法性能。数值实验证明了该改进算法的有效性,并对改进算法的阈值参数进行了率定。

Modified Shuffled Frog Leaping Algorithm based on threshold selection strategy

Shuffled Frog Leaping Algorithm(SFLA) is a new meta-heuristic population evolutionary algorithm and it has fast calculation speed and excellent global search capability.Firstly,the paper introduces the principle of SFLA.Then,aiming at the problem of the individual large space gap between before and after update operation slowering the convergence speed because of the local update strategy,the paper raises a modified shuffled frog leaping algorithm based on the threshold selection strategy.If not satisfying the threshold constraint,the individual element does not update.The modified strategy reduces the individual space gap and improves the capability of the algorithm.Finally,the paper proves validity of the modified algorithm and rates the threshold parameter with numerical experiment.