实现闭环系统参数─致估计的直接辨识方法  

A Direct Identification Method for Consistent Parameter Estimation of Closed-loop Systems

作　　者：张颖[1] 冯纯伯[1] 

机构地区：[1]东南大学自动化研究所,新加坡南洋理工大学

出　　处：《控制与决策》1996年第6期628-632,共5页Control and Decision

基　　金：国家自然科学基金

摘　　要：为了实现闭环系统参数的一致估计，提出了一种一致辨识闭环系统模型参数的偏差补偿最小二乘法。这种方法不需要事先假设系统模型阶次和通道时延满足一定的不等式，允许前向通道和反馈通道之一中的噪声为有色噪声，并且在其辨识过程中不需要对有色噪声建模。bias-eliminated least-squares method is proposed in this paper to obtain consistent parameter estimation of closed-loop systems. It has been shown that in the presented method the identification results are independent of the inequalities essential to other existing closed-loop identification methods. Besides,the noise in one of the forward path and the feedback path is allowed to be colored and need not to be modelled in the identification procedures.

关 键 词：闭环系统 参数估计 直接辨识法 

分 类 号：O231[理学—运筹学与控制论]