约束优化进化算法综述

发布时间：2018-06-12 17:40

本文选题：进化算法 + 约束处理技术　；参考：《软件学报》2017年06期

【摘要】：约束优化进化算法主要研究如何利用进化计算方法求解约束优化问题,是进化计算领城的一个重要研究课题.约束优化问题求解存在约束区域离散、等式约束、非线性约束等挑战,其问题的本质是,如何处理可行解与不可行解的关系才能使得算法更高效.首先介绍了约束优化问题的定义;然后,系统地分析了目前存在的约束优化方法;同时,基于约束处理机制,将这些方法分为罚函数法、可行性法则、随机排序法、?-约束处理法、多目标优化法、混合法等6类,并从约束处理方法的角度对约束优化进化算法的最新研究进展进行综述;最后,指出约束优化进化算法需进一步研究的方向与关键问题.
[Abstract]:The evolutionary algorithm of constrained optimization is an important research topic of evolutionary computing, which mainly studies how to solve the constrained optimization problem by using evolutionary computing method. There are some challenges in solving constrained optimization problems, such as discretization of constraint region, equality constraints, nonlinear constraints, etc. The essence of the problem is how to deal with the relationship between feasible solutions and infeasible solutions in order to make the algorithm more efficient. Firstly, the definition of constraint optimization problem is introduced. Then, the existing constraint optimization methods are systematically analyzed. At the same time, these methods are divided into penalty function method and feasibility method based on constraint processing mechanism. There are six categories of stochastic sorting method, multi-objective optimization method, mixed method and so on, and the latest research progress of constrained optimization evolutionary algorithm is reviewed from the point of view of constraint processing method. The paper points out the direction and key problems of constrained optimization evolutionary algorithm for further study.
【作者单位】：湖南大学信息科学与工程学院;湖南省嵌入式与网络计算重点实验室;
【基金】：国家自然科学基金(61173107,61672215,91320103,61672217) 广东省教育部产学研结合重大科技专项(2012A090300003) 广东省科技计划(2013B090700003) 湖南省研究生科研创新项目(CX2016B067)~~
【分类号】：TP18

【相似文献】