基于对位学习的多目标遗传算法  

Multi-objective Genetic Algorithm Based on Opposition Learning

作　　者：李勇[1] 刘建昌[1] 王昱[2] 

机构地区：[1]东北大学流程工业综合自动化教育部重点实验室,沈阳110006 [2]沈阳航空工业学院自动化学院,辽宁110136

出　　处：《系统仿真学报》2009年第24期7801-7804,共4页Journal of System Simulation

基　　金：高等学校博士学科点专项科研基金(20060145025);辽宁省自然科学基金(20052033)

摘　　要：为了加强非劣排序遗传算法的搜索能力、加快其收敛速度,将对位学习与非劣排序遗传算法相结合,提出了一种用于解决多目标优化问题,基于对位学习的多目标遗传算法,并应用双目标、三目标的测试函数对该算法进行了验证。结果表明,基于对位学习的多目标遗传算法具有较好的收敛性和分布性。In order to enhance search ability and expedite convergence rate of the non-dominated sorting genetic algorithm,a opposition learning based multi-objective genetic algorithm was proposed on the basis of combining opposition based learning and non-dominated sorting genetic algorithm,which aimed at the multi-objective optimization problem.Test functions with two objectives and three objectives were selected to confirm the performance of the proposed algorithm.Simulation results show that the proposed algorithm has better performances both in convergence and distribution.

关 键 词：对位学习 多目标遗传算法 多目标优化 算法验证 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]