Oku, Hiroshi (Osaka Inst. of Tech.), Ogura, Yasuko (Osaka Inst. of Tech.), Fujii, Takao (Fukui Univ. of Tech.)

Closed-Loop Subspace Model Identification Based on QR-Factorization

Scheduled for presentation during the Regular Session "Linear system identification I" (MoP09), Monday, July 24, 2006, 15:20−15:45, Room J

17th International Symposium on Mathematical Theory of Networks and Systems, July 24-28, 2006, Kyoto, Japan

This information is tentative and subject to change. Compiled on April 26, 2024

Abstract

In this paper, a closed-loop subspace model identification method based on QR-factorization is developed. Although external exciting signals and a priori knowledge of a controller in closed-loop are required, the procedure of the proposed method is very similar to the procedures of the existing MOESP methods. Since it can be regarded as a modified version of the method previously proposed by the authors presented at CDC'04, which is derived by integrating PI-MOESP into the Two-stage method, the proposed method can deal with systems contaminated with colored noises. Simulation examples are provided to compare the proposed method with some of the existing methods, that is, CCA, SSARX and PO-MOESP, and demonstrate that the proposed method is capable of identification not only of an ARMAX model but of a Box-Jenkins model.