一种多模态单亲遗传算法

遗传算法在处理复杂的、多模态优化问题时常不十分有效,很难同时搜索多个峰点.这主要是由全局选择机制和交叉算子引起的.针对上述不足,本文提出了一种多模态单亲遗传算法,目标不是发现一个最优解而是多个最优或次优解的集合.主要是对交叉算子和选择机制作了改进,群体中个体能较好地保留自己的遗传特性,大大增强了种群个体的分散性.该方法不仅易实现并行或分布计算,且群体规模可以任意选取.仿真结果验证了算法的有效性.

A MULTIMODAL PARTHENO-GENETIC ALGORITHM

Often a classical GA is not flexible or effective for a optimizing application of complicated multimodal problems. It is difficult to reach more than one of the global maximum, simultaneously. Considering the above deficiency of GA, the paper presents a multimodal partheno genetic algorithm (MPGA). The aim of the proposed algorithm is to find not only a single solution but also a set of the optimal. Both new a crossover operator and selection schemes are introduced so that every population individual inherits more genetic material of its own forefather. The diversity of population individual can be improved greatly in this approach. The size of population is very flexible and the architecture of computation is well suited for parallel implementation. The simulating cases show the effectiveness of this approach.