多变量系统高效耦合递推最小二乘辨识算法及计算量分析

Highly-efficient coupled recursive least squares identification algorithm for multivariable systems and its computational amount analysis

中文关键词  参数估计  耦合辨识概念  递推辨识  最小二乘  多变量系统

英文关键词  parameter estimation  coupling identification concept  recursive identification  least squares  multivariable systems

中文摘要

      针对多变量系统维数大、参数数目多, 导致辨识算法计算量大的问题, 基于耦合辨识概念, 推导了多变量系统的高效耦合递推最小二乘算法. 算法的主要思想是根据各子系统辨识模型的特点, 通过耦合辨识概念将子系统间相同的参数向量耦合起来, 避免子系统参数向量的冗余估计. 通过算法计算量分析表明, 与递推最小二乘算法相比, 所提出算法具有较少的计算量. 仿真实例验证了算法的有效性.

英文摘要

      Because of the a large-scale multivariable system has a large number of parameters and its identification algorithms require a large amount of computation, a highly-efficient coupled recursive least squares algorithm is derived for a multivariable system based on the coupling identification concept. The main idea of the algorithm is to couple the same parameter vectors among subsystems according to the characteristics of each subsystem identification model, so as to avoid the redundant estimation of the subsystem parameter vectors. The computational efficiency analysis shows that the proposed algorithm has less amount of computation than the recursive least squares algorithm. The simulation example verifies the effectiveness of the proposed algorithm.