Improved Closed-Loop Subspace Identification with Prior Information Using Principal Component Analysis and Column Weighting

It has been proven that integrating prior information into subspace identification methods (SIMs) can improve the accuracy of model identification. In this study, a new closed-loop SIM integrating prior information is proposed based on principal component analysis (PCA) and column weighting. Compared with the conventional SIMs based on PCA, the constrained least square (CLS) approach is used to incorporate prior information into the impulse response after the PCA procedure. Then the system matrices can be extracted from the estimated impulse response parameters. The simulation results demonstrate that the proposed method is more accurate and stable for model identification.