Abstract: To improve the performance of particle swarm optimization algorithm, the idea of “catfish effect” is introduced to transform the individual evolutionary strategies of particle swarm. Chaos method is used to improve the population search strategy, and the two are put together, not only improving population breadth search ability, but also enhancing the depth of search capability. With the differential evolution algorithm to be mixed, algorithms complement each other, forming a novel hybrid algorithm, better coordinating the contradiction between the breadth search and depth search, and enhancing the performance of the algorithm. After testing the three standard functions, the simulation results show that this algorithm’s capability of escaping from local trap and search accuracy are significantly improved.

Key words: particle swarm optimization algorithm, catfish effect, chaos, differential evolution algorithm, hybrid algorithm

摘要： 为了改善粒子群优化算法的性能，引入了“鲶鱼效应”思想，改造粒子群个体的进化策略，用混沌方法改良了种群搜索策略，把这两者结合起来，既提高种群的广度搜索能力，又提升深度搜索能力，跟差分进化算法进行混合，算法优势互补，形成一种新型的混合算法，更好地协调广度搜索和深度搜索之间的矛盾，提升算法性能。经过对三个标准函数的测试，仿真结果表明该算法在逃离局部陷阱能力和搜索精度均有显著提高。

关键词: 粒子群优化算法, 鲶鱼效应, 混沌, 差分进化算法, 混合算法