中南大学学报（自然科学版）2012，Vol.43Issue(9)：3478-3484,7.
基于适应性均衡模型的CMAES约束优化算法
Constrained evolutionary optimization of CMAES based on adaptive tradeoff model
摘要
Abstract
A novel hybrid algorithm named ATM-CMAES, which integrated covariance matrix adaptation evolution strategy (CMAES) with adaptive tradeoff model (ATM), was proposed to solve constrained optimization problems. CMAES in the new algorithm used the relationship between the best subgroup of the current generation and the mean of the previous generation to update covariance matrix. Compared to traditional evolutionary algorithm, CMAES needed smaller evolving population size since it combined rank-1 with rank-μ update mechanism in the update process of covariance matrix. ATM was introduced as constraint-handing technique in ATM-CMAES. Three issues, i.e., infeasibility, semi-feasibility and feasibility, were addressed by designing different tradeoff schemes during different stages of a search process to obtain an appropriate tradeoff between objective function and constraint violations. The hybrid algorithm was tested on 13 well-known benchmark functions. The results show that ATM-CMAES is a general, robust and effective method.关键词
约束优化问题/适应性均衡模型/协方差矩阵自适应进化策略/进化算法Key words
constrained optimization problem/adaptive tradeoff model (ATM)/covariance matrix adaptation evolution strategy (CMAES)/evolutionary algorithm分类
信息技术与安全科学引用本文复制引用
黄亚飞,梁昔明,陈义雄..基于适应性均衡模型的CMAES约束优化算法[J].中南大学学报（自然科学版）,2012,43(9):3478-3484,7.基金项目
国家自然科学基金资助项目(60874070) （60874070）
国家留学回国人员科研启动基金资助项目([2008]375) （[2008]375）
湖南省教育厅项目(10C0373) （10C0373）
