基于Pareto-ε优胜的自适应快速多目标演化算法  

Quick multi-objective evolutionary algorithm based on adaptive Pareto-ε dominance

作　　者：王江晴[1] 杨勋[2] 

机构地区：[1]中南民族大学计算机科学学院,武汉430074 [2]中国延安干部学院办公厅,陕西延安716000

出　　处：《计算机应用》2010年第4期997-999,共3页journal of Computer Applications

基　　金：国家自然科学基金资助项目(60842004)

摘　　要：在多目标优化领域,如何快速地为决策者提供合理、可行的解决方案尤为重要,为此,给出了多目标优化问题的一种新解法。定义了一种Pareto-ε优胜关系的概念,将此概念引入多目标优化问题中,设计了一种新的基于ε-优胜的自适应快速多目标演化算法。计算机仿真表明,该算法可以明显改善求解多目标优化问题时的寻优过程,能适应实际应用环境下快速、有效的决策要求。For Multi-objective Optimization Problems（MOP）,it is very important to provide proper and feasible solutions rapidly for the decision makers.A method for MOP was given.First,a concept of Pareto-ε dominance was defined.Then,based on this concept,a new adaptive multi-objective evolutionary algorithm was proposed.The simulation results demonstrate that the new algorithm can improve the process of MOP optimization,and can meet the requirements of high-speed and effectiveness in application.

关 键 词：多目标优化 Pareto优胜 PARETO前沿 演化算法 自适应 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]