Input selection in N2SID using group lasso regularization

Input selection in N2SID using group lasso regularization

Klingspor, M. (Linköping University)
Hansson, A (Linköping University)
Löfberg, J. (Linköping University)
Verhaegen, M.H.G. (TU Delft Team Raf Van de Plas)

Dochain, Denis (editor)
Henrion, Didier (editor)
Peaucelle, Dimitri (editor)

2017

Input selection is an important and oftentimes difficult challenge in system identification. In order to achieve less complex models, irrelevant inputs should be methodically and correctly discarded before or under the estimation process. In this paper we introduce a novel method of input selection that is carried out as a natural extension in a subspace method. We show that the method robustly and accurately performs input selection at various noise levels and that it provides good model estimates.

Input selection
System identification
State-space models
N2SID
Subspace methods
Signal-to-noise ratio

http://resolver.tudelft.nl/uuid:2fb58095-5c7b-4699-a491-bfd5d1fdb8c1

https://doi.org/10.1016/j.ifacol.2017.08.1472

Elsevier, Laxenburg, Austria

IFAC-PapersOnLine: Proceedings 20th IFAC World Congress, 50-1

20th World Congress of the International Federation of Automatic Control (IFAC), 2017, 2017-07-09 → 2017-07-14, Toulouse, France

IFAC-PapersOnline, 2405-8963, 50 (1)

Institutional Repository

conference paper

© 2017 M. Klingspor, A Hansson, J. Löfberg, M.H.G. Verhaegen