MLGA: A Multilevel Cooperative Genetic Algorithm

摘要

This paper incorporate the multilevel selection (MLS) theory into the genetic algorithm. Based on this theory, a Multilevel Cooperative Genetic Algorithm (MLGA) is presented. In MLGA, a species is subdivided in a set of populations, each population is subdivided in groups, and evolution occurs at two levels so called individual and group level. A fast population dynamics occurs at individual level. At this level, selection occurs between individuals of the same group. The popular genetic operators such as mutation and crossover are applied within groups. A slow population dynamics occurs at group level. At this level, selection occurs between groups of a population. A group level operator so called colonization is applied between groups in which a group is selected as extinct, and replaced by offspring of a colonist group. We used a set of well known numerical functions in order to evaluate performance of the proposed algorithm. The results showed that the MLGA is robust, and provides an efficient way for numerical function optimization.