变量带误差系统的递推辨识:从线性到非线性系统  

RECURSIVE IDENTIFICATION FOR ERRORS-IN-VARIABLES SYSTEMS:FROM LINEAR TO NONLINEAR SYSTEMS

作　　者：陈翰馥[1] 牟必强[1] 

机构地区：[1]中国科学院数学与系统科学研究院系统科学研究所系统控制重点实验室,北京100190

出　　处：《系统科学与数学》2012年第10期1180-1192,共13页Journal of Systems Science and Mathematical Sciences

基　　金：国家自然科学基金项目(61120106011;61134013;61273193);国家数学与交叉科学中心资助课题

摘　　要：介绍变量带误差(EIV)系统的递推辨识方法.在引言中扼要介绍了EIV系统辨识的现状后,分别对多变量线性EIV系统及EIV Hammerstein系统给出了递推辨识算法,并给出条件使这些估计以概率1收敛到真值.最后提出了一些值得进一步研究的问题.The recursive identification methods for errors-in-variables （EIV） systems are described in the paper. After briefly introducing the current status of identification for EIV systems, the recursive identification algorithms for the multivariate linear EIV system as well as for the EIV Hammerstein system are presented. The conditions guaranteeing the strong consistency of the estimates given by the algorithms are demonstrated. Finally, some remarks on problems for future research are pointed out.

关 键 词：变量带误差系统 递推估计 随机逼近 强一致性 

分 类 号：N945.14[自然科学总论—系统科学]