14th IFAC Symposium on System Identification, SYSID 2006

SYSID-2006 Paper Abstract

Close

Paper WeB1.5

Borges, JosÚ (Tech. Univ. of Lisbon, Inst. Superior TÚcnico), Verhaegen, Michel (Delft Univ. of Tech.), Verdult, Vincent (Delft Univ. of Tech.)

Iterative Subspace Identification of Piecewise Linear Systems

Scheduled for presentation during the Regular Session "Identification of Hybrid and Parameter Varying Systems" (WeB1), Wednesday, March 29, 2006, 16:50−17:10, Concert Hall

14th IFAC Symposium on System Identification, March 29 - 31, 2006, Newcastle, Australia

This information is tentative and subject to change. Compiled on July 17, 2018

Keywords Subspace Methods, Recursive Identification, Multivariable System Identification

Abstract

In this paper an iterative subspace approach is proposed for the identification of piecewise linear systems with unknown number of local systems. The proposed approach runs offline and integrates the detection of switching between local systems and the iterative estimation of the respective observability spaces in a single framework - previous approaches treated this problem in a separated way. The parameters of each local linear system, as well as the necessary similarity transformations to bring all local systems to the same basis, are then computed in an efficient way. Monte-Carlo simulations are used to show the effectiveness of the proposed approach.