中文摘要：

常规基于记忆的进化算法在动态环境中往往达不到期望的效果,这主要是由于记忆体大小的限制.为此提出了动态环境下基于可变记忆的进化算法（IMEEA）,其核心思想是算法中拥有两个种群,即搜索种群和记忆种群,同时采用过度变异策略来增加种群的多样性.算法中的两个种群有最小和最大的允许长度,并且种群的大小根据进化过程的进行而不断变化.仿真结果表明,在动态环境中IMEEA算法的跟踪误差要小于常规的记忆提高进化算法（MEEA）,从而证明了所提算法的有效性.

英文摘要：

Traditional memory-based evolutionary algorithms often may not achieve the desired performances in dynamic environments,which is mainly due to the fixed memory size.A variable size memory-based evolutionary algorithm is proposed.The improved memory enhanced evolutionary algorithm（IMEEA）,which combines memory population and search population,and hyper-mutation is used to promote and maintain diversity.The two populations have minimum and maximum sizes allowed that change according to the stage of the evolutionary process.Simulation results show that the tracking error of the IMEEA is less than the memory enhanced evolutionary algorithm（MEEA）,and then prove the effectiveness of this new algorithm.