具有局部定向繁殖策略的自学习进化算法

Active Evolutionary Algorithm with Local-directional Reproduction Strategy

提出一种具有局部定向繁殖策略的自学习进化算法(Active Evolutionary Algorithm with Local-directional ReproductionStrategy,LRSEA),它在每一代通过进化个体之间适应度差异找到最优进化方向,然后在此方向附近用局部定向繁殖策略进行局部搜索.针对目前进化算法研究中存在对个体主观能动性考虑不足,没有充分利用进化中"优秀信息"的问题,将主动进化机制引入到该算法中,提出反映个体学习能力的积极学习算子.实验结果表明:该算法收敛速度和局部搜索能力有明显改善.

This paper presents an Active Evolutionary Algorithm with Local-directional Reproduction Strategy,which uses the difference of fitness that between generations to find the optimal evolution direction each generation,and local-directional reproduction strategy based local search is executed nearby the optimal evolution direction.And considering the main problems of traditional evolutionary algorithms which do not make full use of the individual subjective initiative and the 'excellent information',an active evolutionary mechanism is imposed into them and an active learning operator is proposed.Experiment results show that this new algorithm increases the speed of convergence and enhances the ability of local search.