针对标准遗传算法中的早熟收敛现象,提出一种基于竞争协同进化的改进遗传算法.该算法根据个体与对手竞争的表现来衡量个体的生存能力,生存能力由个体所击败对手的数量和优秀程度决定,个体在击败更多更优对手的努力中逐步进化.函数优化实验结果表明,该算法收敛速度快,且能有效保留种群多样性,与标准遗传算法及其他多种群遗传算法相比,能有效减轻早熟收敛现象.
To solve the problem of premature convergence in standard genetic algorithms(SGAs),this paper presents a modified genetic algorithm based on competitive coevolution(MGACC),in which an individual survival ability was assessed by its competitive performance relative to its opponents.The individual survival ability was determined by the number and characteristic of the opponents it defeats.All individuals were refined gradually during the endeavor to defeat more opponents or excellent opponents.Experimental re...