一种基于密度距离的多目标优化问题求解方法  

A Multi-objective Optimization Problem Solving Method Based on Density Distance

作　　者：胡俊 张平华 程晨[1] HU Jun;ZHANG Ping-hua;CHENG Chen(Information Engineering College, Anhui Institute of International Business, Hefei Anhui231131, China;Information Engineering and Media College, Hefei Technology College, Hefei Anhui231131, China)

机构地区：[1]安徽国际商务职业学院信息工程学院,安徽合肥231131 [2]合肥职业技术学院信息工程与传媒学院,安徽合肥231131

出　　处：《菏泽学院学报》2019年第2期21-24,共4页Journal of Heze University

基　　金：2016安徽省教育厅自然科学研究重点项目(KJ2016A304);2017安徽省教育厅自然科学研究重点项目(KJ2017A624);安徽省高校自然科学重点项目(KJ2018A0827)

摘　　要：多目标进化算法是一种具有广泛的应用领域和应用前景的算法,它由适应度赋值和外部种群维护两个组成部分,本文提出了一种基于个体密度距离的外部种群方法,进而根据Pareto支配关系,提出用个体密度距离定义的一种适应度函数,实现以个体密度距离为基础的多目标算法优化.Multi-objective evolutionary algorithm (MOEA) is an algorithm with a wide range of applications and prospects. It consists of fitness assignment and external population maintenance. In this paper, an external population method based on individual density distance is proposed. According to Pareto dominance relationship, an fitness function is defined to realize multi-objective optimization based on individual density distance.

关 键 词：密集距离 多目标进化 算法研究 适应度赋值 外部种群维护 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]