摘要
简单讨论了行列式、矩阵逆和块矩阵逆的计算量;研究了信息向量耦合型多变量系统的子系统递推最小二乘辨识方法,给出了计算量小的联合递推最小二乘辨识算法;研究了部分信息向量耦合型多变量系统的子系统最小二乘辨识算法,提出了计算量小的基于块矩阵求逆的最小二乘辨识算法;给出了部分信息向量耦合型多变量系统的子系统递推最小二乘辨识算法,提出和推导了基于辨识模型分解的递推最小二乘辨识算法,并分别讨论了提出算法的计算量.
This paper discusses the computational efficiency of the determinant, the matrix inversion and the block matrix inversion. For the multivariable systems with coupled information vectors, we study the subsystem recursive least squares （RLS） identification algorithm and the joint RLS identification agorithm with less computation are studied. For the multivariable systems the partially coupled information vectors, the subsystem least squares identifi- cation algorithm and the block matrix inversion based least squares identification algorithm are presented and the subsystem RLS identification algorithm and the identification model decomposition based RLS identification algo- rithm are proposed. Finally, the eomptutational efficiency of the proposed algorithms is analyzed.
出处
《南京信息工程大学学报（自然科学版）》
CAS
2012年第6期481-495,共15页
Journal of Nanjing University of Information Science & Technology（Natural Science Edition）
基金
国家自然科学基金(61273194)
江苏省自然科学基金(BK2012549)
高等学校学科创新引智计划(B12018)
