动态环境下基于聚类的克隆选择算法

A Cluster-based Clonal Selection Algorithm in Dynamic Environment

提出一种在动态环境下基于聚类的克隆选择算法.在进化过程中,首先根据各个个体的空间位置,将整个种群划分成不同的聚类;其次,各个聚类共享信息,促进每个聚类各自进化.在每个聚类内部引入学习策略和交互策略来提高个体的搜索能力.另外,把每次找到的潜在的局部最优个体存储起来,在这些个体上运用局部搜索算子可以有效的提高搜索效果.在环境发生改变之后,被存储起来的潜在的最优个体可以用于最优目标的跟踪,从而节省计算资源.

A cluster-based clonal selection algorithm in dynamic environment was presented.The population was partitioned into multiple clusters according to the spatial location at first,and then each cluster evolved separately by sharing information.The learning strategy within the cluster and interaction among clusters were introduced to the hypermutation to improve search ability.In addition,the possible optima were recorded as memory,the searching ability could be improved when the local operator on these potential local optimal individuals was applied.After the environment has been changed,the stored potential local optimal individuals can be used to tracking optima,as a result,computed result can be saved.