摘要
在单目标元胞遗传算法的基础上,提出一种多目标元胞遗传算法(Multi-Objective Cellular Genetic Algorithm,MOC-GA)。该算法使用元胞自动机的生命游戏规则替代遗传算法的交叉算子,使用NSGA-Ⅱ的选择方式选择个体,同时设置外部种群存储算法进化过程中的非支配个体。为了提高算法的效率和保证非支配解集良好的分布性,使用改进的快速排序法选择非支配个体,依照个体的动态聚集距离对外部种群进行消减。与NSGA-Ⅱ相比,实例表明,该算法具有更好的收敛性和稳定性。
Proposed a Multi-Objective Cellular Genetic Algorithm(MOCGA),which was based on the cellular genetic algorithm.The proposed algorithm replaced the crossover operator of genetic algorithm by the rule of"the game of life",used the selection method of NSGA-II and a external population to store the non-dominated individual and used dynamic crowding distance to generate a better distribution of solutions along the front.The new algorithm was tested for comparing the performance of NSGA-II.The results show the better convergence and solution diversity of new algorithm.
出处
《现代制造工程》
CSCD
北大核心
2010年第7期46-50,共5页
Modern Manufacturing Engineering
关键词
元胞自动机
遗传算法
多目标优化
Cellular Automata(CA)
Genetic Algorithm(GA)
multi-objectives optimization