基于进化算法的鲁棒控制研究进展  

Survey of robust control based on evolutionary algorithm

作　　者：杨智民[1] 郑言海[1] 庄显义[1] 

机构地区：[1]哈尔滨工业大学,黑龙江哈尔滨150001

出　　处：《电机与控制学报》1999年第4期235-240,共6页Electric Machines and Control

摘　　要：进化算法是增减自然遗传学机理基础上的参数搜索方法，本文综述了近年来基于进化算法的鲁棒控制系统设计方面的研究情况。首先简单介绍了进化算法及其改进，讨论了进化算法以及传统优化方法的优点和不足；然后阐述了如何将进化算法应用到不同的鲁棒方法当中，论述了进化算法在控制系统鲁棒性设计，同时镇定控制、Ｈ＾∞次优控制器设计、Ｈ＾∞权函数选择、特征结构配置、μ综合理论以及定量反馈理论中的应用成果；Evolutionary algorithms are parameter search procedures based on nature genetic mechanism. In this paper, a survey of robust control based on the evolutionary algorithm in recent years is presented. Firstly, evolutionary algorithm and its improvements are introduced briefly. Both merits and drawbacks of genetic algorithm and traditional optimizing methods are discussed. Then how to apply evolutionary algorithm to various robust design methods is discussed. A range of applications in robust control system design, simultaneous stabilization, H∞ sub-optimal controller design, H∞ weight function selection, eigen-structure placement, μ synthesis and quantitative feedback control theory are reviewed. Finally, further research field is presented.

关 键 词：进化算法 鲁棒控制 控制理论 H^∞控制器 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] O231[自动化与计算机技术—控制科学与工程]