Using Engineering View to Design Better Multi-objective Evolutionary Algorithms
首发时间:2006-11-30
Abstract:Multi-Objective Optimization Problems(MOPs) are very difficult to solve by traditional methods, so many Evolutionary Algorithms were designed to deal with them. In this paper, we discussed Multi-objective Evolutionary Algorithms(MOEAs) in engineering view, and proposed a new unified framework. According to the framework, we find some possible key points to the design of MOEAs and these key points are discussed. Then we introduce a new MOEA and test it with some famous benchmark functions. The numerical experiments show that the algorithm can obtain more non-dominant solutions which distribute equably and approximately to the Pareto front in less time than some algorithms such as SPEA2, NSGAII and HPMOEA.
keywords: Multi-Objective Optimization, Framework, Evolutionary Algorithm, Pareto Front
点击查看论文中文信息
Using Engineering View to Design Better Multi-objective Evolutionary Algorithms
摘要:Multi-Objective Optimization Problems(MOPs) are very difficult to solve by traditional methods, so many Evolutionary Algorithms were designed to deal with them. In this paper, we discussed Multi-objective Evolutionary Algorithms(MOEAs) in engineering view, and proposed a new unified framework. According to the framework, we find some possible key points to the design of MOEAs and these key points are discussed. Then we introduce a new MOEA and test it with some famous benchmark functions. The numerical experiments show that the algorithm can obtain more non-dominant solutions which distribute equably and approximately to the Pareto front in less time than some algorithms such as SPEA2, NSGAII and HPMOEA.
关键词: Multi-Objective Optimization, Framework, Evolutionary Algorithm, Pareto Front
论文图表:
引用
No.1008889298116488****
同行评议
勘误表
Using Engineering View to Design Better Multi-objective Evolutionary Algorithms
评论
全部评论0/1000