In order to overcome the deficiency of NSCA II on distribution, this paper proposed a multi-objective co-evolutionary algorithm based on culture. Hie algorithm proposed an index which was used for evaluating the diversity of belief space, extracted knowledge from belief space, and utilized the knowledge to guide the evolution of population. It proposed a concept called knowledge hypothesis set, created new knowledges based on knowledges in being to enhance local search in sparse area, which speeded up the convergence of the algorithm. Simulation results indicate that the algorithm improves signifi-cantly in terms of the ability of convergence and distribution.%为了改进NSGAⅡ算法中存在的分布性等不理想问题,在NSGAⅡ的基础上提出了基于文化的多目标协同进化算法.该算法提出评测信念空间多样性的指标,从信念空间中提取知识,利用知识来指导种群的进化;提出知识假说集,以现有知识为基础产生新知识,加强局部搜索,加速算法收敛.仿真实验表明该算法较NSGAⅡ在收敛性及分布性方面均有明显提高.
展开▼