论文部分内容阅读
提出一种基于辅助变量的子空间辨识方法,适用于控制器信息未知以及参考输入已知的闭环系统参数辨识.通过将输入-输出数据块正交投影到辅助变量的行空间,直接得到扩展观测矩阵垂空间的估计.由此可从闭环系统中提取出对象模型信息,同时由SVD分解得到扩展观测矩阵与下三角Toeplitz矩阵的估计.给出了系统参数矩阵、噪声矩阵的计算方法.将所提出的子空间辨识方法应用于闭环动态的系统参数估计,其结果表明了该方法的有效性.
A subspace identification method based on auxiliary variables is proposed, which is suitable for parameter identification of closed-loop system with unknown controller information and known reference input. By directly projecting input-output data blocks into the row space of auxiliary variables, extended observation Matrix vertical space, the information of the object model can be extracted from the closed-loop system, and the estimation of the extended observation matrix and the lower triangular Toeplitz matrix is obtained by SVD decomposition. The calculation method of system parameter matrix and noise matrix is given. The proposed subspace identification method is applied to the estimation of closed-loop dynamic system parameters. The results show the effectiveness of the proposed method.