Convex Relaxation Approach to the Identification of the Wiener-Hammerstein Model
2008 (English)In: Proceedings of the 47th IEEE Conference on Decision and Control, 2008, 1375-1382 p.Conference paper (Refereed)
In this paper, an input/output system identificationtechnique for the Wiener-Hammerstein model and its feedbackextension is proposed. In the proposed framework, the identificationof the nonlinearity is non-parametric. The identificationproblem can be formulated as a non-convex quadratic program(QP). A convex semidefinite programming (SDP) relaxation isthen formulated and solved to obtain a sub-optimal solution tothe original non-convex QP. The convex relaxation turns out tobe tight in most cases. Combined with the use of local search,high quality solutions to the Wiener-Hammerstein identificationcan frequently be found. As an application example, randomlygenerated Wiener-Hammerstein models are identified.
Place, publisher, year, edition, pages
2008. 1375-1382 p.
IdentifiersURN: urn:nbn:se:kth:diva-82410DOI: 10.1109/CDC.2008.4739417OAI: oai:DiVA.org:kth-82410DiVA: diva2:498212
47th IEEE Conference on Decision and Control, 2008, CDC 2008. Cancun, Mexico. Dec. 9-11, 2008
QC 201205142012-02-112012-02-112012-05-14Bibliographically approved