动态优化算法综述

动态优化算法综述 Survey on Dynamic Optimization Algorithms 陈莉 CHEN Li1 1 2 first-author 丁立新 DING Lixin1 1 武汉大学软件工程国家重点实验室 武汉大学软件工程国家重点实验室 State Key Laboratory of Software Engineering,Wuhan University State Key Laboratory of Software Engineering,Wuhan University 空军雷达学院预警监视情报系 空军雷达学院预警监视情报系 Department of Early Warning Surveillance Intelligence,Air Force Radar Academy Department of Early Warning Surveillance Intelligence,Air Force Radar Academy 动态优化算法的研究已成为优化算法领域研究的一个热点.对于基于种群的优化算法而言,它主要可以分为环境变化后增加多样性的方法、运行过程中始终保持多样性的方法、基于记忆机制的方法、多种群方法和基于预测机制方法5类.动态优化算法的关键是在搜索过程中始终保持搜索空间开发和探索之间平衡.该类算法不仅能发现最优个体,而且能在动态环境中跟踪变化了的最优个体.在今后的动态优化研究中,重点应放在动态优化算法理论方面和算法设计、构建上,使它更接近现实问题. Research on Algorithms dealing with dynamic optimization problems has been one of the hotspots in the optimization algorithms’ area.The population-based optimization algorithms are grouped into five categories: increasing diversity after environment changes,keeping diversity during the run,using memory schemes,multi-population and prediction-based approaches.The keys of these methods are keeping the balance of the exploration and exploitation in research space.The algorithms can not only find the optimum but also track the changing optimum.At last this paper points out the problems in the dynamic optimization area needed to research deeply in the future.Dynamic optimization algorithms need researches on algorithm design,building near-real-world dynamic optimization problem model and algorithm theory for dynamic optimization in the future. 动态优化 多样性 多种群 记忆机制 预测机制 dynamic optimization diversity multi-population memory schemes prediction 10.14188/j.1671-8836.2011.03.009 TP301.6 国家自然科学基金（60975050）;高等学校博士点基金（20070486081）;湖北省杰出青年基金（2005ABB017）资助项目 2011-03-01 2021-04-01 3 57