基于微粒群与混合蛙跳融合的群体智能算法

针对微粒群算法和混合蛙跳算法存在的早熟收敛问题，提出一种基于微粒群与混合蛙跳算法融合的群体智能算法。新算法将整个群体分成数目相等的蛙群和微粒群群体。在两群体独立进化过程中，设计了一种两群之间的信息替换策略：比较蛙群与微粒群的最佳适应值，如果蛙群进化较好，利用蛙群各子群中最差个体替换微粒群一部分较好个体；否则，用微粒群中较好的一部分个体替换蛙群各子群的最好个体。同时，设计了一种两群之间的相互协作方式。为避免微粒群因早熟收敛而影响信息替换策略效果，适时对其所有个体最好位置进行随机扰动。仿真实验表明，新算法可以有效提高全局搜索能力及收敛速度，对于高维复杂函数问题，算法具有很好的稳定性。

Swarm intelligence algorithm based on combination of shuffled frog leaping algorithm and particle swarm optimization

Concerning the premature convergence of Particle Swarm Optimization(PSO) algorithm and Shuffled Frog Leaping Algorithm(SFLA),this paper proposed a swarm intelligence optimization algorithm based on the combination of SFLA and PSO.In this algorithm,the whole particle was divided into two equal groups: SFLA and PSO.An information replacement strategy was designed in the process of their iteration: comparing the fitness of PSO with that of SFLA,the worst individual in each subgroup of SFLA would replace some better individuals in PSO when SFLA is better;otherwise,some better individuals in PSO would replace the best individual in each subgroup of SFLA.Meanwhile,a collaborative approach between the two groups was also designed.Since the information replacement strategy could be influenced by the premature convergence problem in PSO,a random disturbance would be given on each particle’s best position.The simulation results show that the proposed algorithm can improve the global search ability and convergence speed efficiently.For the complex functions with high-dimension,the algorithm has very good stability.