一种针对高维决策空间的进化多目标优化方法  

Evolutionary Multiobjective Optimization Method for High Dimensional Decision Space

作　　者：向勇[1] 唐常杰[1] 曾涛[2] 朱明放[1] 邱江涛[1] 王蓉[3] 

机构地区：[1]四川大学计算机学院,四川成都610065 [2]天津师范大学计算机与信息工程学院,天津300387 [3]成都电子机械高等专科学校计算机工程系,四川成都610031

出　　处：《系统仿真学报》2008年第9期2329-2333,共5页Journal of System Simulation

基　　金：国家自然科学基金资助项目(60773169);四川省教育厅资助科研项目(2006B067);成都电子机械高等专科学校科研项目(KY061007B);天津师范大学引进人才基金项目(5RL062)

摘　　要：进化算法可并行处理多个解的特性使得它特别适合解决多目标优化问题。针对高维决策空间,将基因表达式编程引入多目标优化,设计了新的个体结构和操作,提出了一个进化多目标优化算法EMOGEP。实验结果表明,新算法在低维决策空间是可行和有效的;在高维决策空间中,表现出了比传统进化多目标优化算法更好的性能;多模态情况下,新算法能很好的逼近理论Pareto前沿。Evolutionary algorithms are particularly suited for Multiobjective Optimization Problem （MOP）, because they process a set of solutions in parallel. Aiming at high dimensional decision space, an innovative algorithm named Evolutionary Multiobjective Optimization Gene Expression Programming （EMOGEP） was presented. A new structure for individual was designed and some genetic operators were proposed. Experiment results show that EMOGEP is feasible and effective in low dimensional decision space. Especially in high dimension, EMOGEP performs better than traditional EMO algorithms. In the circumstance of multimodality, EMOGEP approximately converges to the true Pareto-optimal front.

关 键 词：高维 多目标优化 进化算法 基因表达式编程 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] TP18[自动化与计算机技术—计算机科学与技术]