Closed-Loop Subspace Identification Algorithm of EIV Model Based on Orthogonal Decomposition and PCA
In this paper, after analysis of the reason why some existing subspace methods may deliver a bias in the closed-loop conditions, a new SIM for closed-loop system based on orthogonal decomposition and principal component analysis is proposed by adopting the EIV model structure. Then, the underlying reason why SIMPCA-Wc delivers a bias estimate is explained from realization theory of closed-loop system based on orthogonal decomposition. At last, simulations show that the proposed method ORT_PCA-Wc used for closed-loop EIV system is effective and feasible.
Keywordssubspace identification methods (SIMs) closed-loop identification orthogonal decomposition principal component analysis (PCA) error in variable (EIV)
Unable to display preview. Download preview PDF.
- 2.Jansson, M.: Subspace identification and ARX modeling. In: Proceedings of the 13th IFAC SYSID Symposium (2003)Google Scholar
- 3.Qin, S.J., Ljung, L.: Closed-loop subspace identification with innovation estimation. In: Proceedings of SYSID 2003, Rotterdam (2003)Google Scholar
- 11.Wang, J., Qin, S.J.: Principal component analysis for errors-in-variables subspace identification. In: Proceedings of the 40th IEEE Conference on Decision and Control, Orlando, Florida, pp. 3934–3941 (2001)Google Scholar