一种用于非参数系统辨识的粗糙进化方法(英文)  

A Rough Evolutionary Approach to Non Parametric System Identification

作　　者：刘健勤[1] 刘娟[1] 蔡自兴[1] 

机构地区：[1]中南工业大学信息工程学院,长沙410083

出　　处：《控制理论与应用》2000年第3期405-409,共5页Control Theory & Applications

基　　金：Foundationitem :supportedbytheNationalNaturalScienceFoundationofChina (699740 4 3)andtheNaturalScienceFoundationofHunanProvince(

摘　　要：提出了一种用于非参数系统辨识的粗糙进化方法 .该方法由嵌入ApEn的进化计算 (EC)、进化粗糙集、基于信息论的组合型进化策略 (ES)和鲁棒非参数系统辨识所构成 .该方法已用于Logistic方程的辨识 ,这对于诸如家用电器的实时系统来说也属于一种实用的技术 .This paper proposes a rough evolutionary approach to non parametric system identification. The approach consists of EC(Evolutionary Computation) embedded with ApEn, evolutionary rough set, combinatorial ES (Evolutionary Strategy) based on information theory and robust non parametric identification. This method has been used to identify the Logistic equation and it has been proven a useful technique in real time system, especially household electric appliances. The result of the experiment shows that our method is practical and efficient.

关 键 词：进化计算 粗糙集 非参数系统辨识 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]