A New Subspace Identification Method for Open and Closed Loop Data
2005 (English)In: IFAC Proceedings Volumes (IFAC-PapersOnline): Volume 16, 2005, 2005, 500-505 p.Conference paper (Refereed)
Abstract: Subspace methods have emerged as useful tools for the identiﬁcation of lineartime invariant discrete time systems. Most of the methods have been developed for theopen loop case to avoid difﬁculties with data correlations due to the feedback. This paperextends some recent ideas for developing subspace methods that can perform well on datacollected both in open and closed loop conditions. Here, a method that aims at minimizingthe prediction errors in several approximate steps is proposed. The steps involve usingconstrained least squares estimation on models with different degrees of structure such asblock-toeplitz, and reduced rank matrices. The statistical estimation performance of themethod is shown to be competitive to existing subspace methods in a simulation example.
Place, publisher, year, edition, pages
2005. 500-505 p.
multivariable, identiﬁcation, Subspace methods
Control Engineering Signal Processing
IdentifiersURN: urn:nbn:se:kth:diva-82577ScopusID: 2-s2.0-79960746761ISBN: 008045108XISBN: 978-008045108-4OAI: oai:DiVA.org:kth-82577DiVA: diva2:498381
16th IFAC World Congress, Prague, Czech Republic, Jul. 4-8, 2005
QC 201202192012-02-122012-02-122012-10-01Bibliographically approved