引用本文:宋文忠 ,徐嗣鑫.多变量线性系统结构和参数辨识的递推最小二乘法[J].控制理论与应用,1984,1(4):64~73.[点击复制]
Song Wenzhong,Xu Sixin.RECURSIVE LEAST SQUARES ALGORITHM FOR STRUCTURAL AND PARAMETRIC IDENTIFICATION OF MULTIVARIABLE SYSTEMS[J].Control Theory and Technology,1984,1(4):64~73.[点击复制]
多变量线性系统结构和参数辨识的递推最小二乘法
RECURSIVE LEAST SQUARES ALGORITHM FOR STRUCTURAL AND PARAMETRIC IDENTIFICATION OF MULTIVARIABLE SYSTEMS
摘要点击 827  全文点击 383  投稿时间:1983-04-10  修订日期:1983-09-03
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  
  1984,1(4):64-73
中文关键词  
英文关键词  
基金项目  
作者单位
宋文忠 ,徐嗣鑫 南京工学院 
中文摘要
      本文研究多变量线性系统的结果和参数辨识,利用解不定对称系数方程组的方法,导出了搜索系统结构特征值的最小二乘递推算法。并提出了一个判断系统结构的新准则——系统特征方程系数符号检验法,把两者结合在一起,构成了完整的多变量系统结构和参数辨识的递推最小二乘法,并给出了仿真例子。 本方法也可作为最小二乘的各种改进算法(如广义最小二乘法等)的基础。
英文摘要
      This paper deals with the structural and parametric identification of multivariable systems. A new recursive least squares algorithm to search for system structural parameters using the solution of undetermined linear equations with symmetric coefficients, and a criterion to test the system structure using the signs of coefficients of the system characteristics equation are developed. A simulated example is given. This recursive algorithm is also applicable to various modified least squares methods, such as the extended least squares method etc..