On Model Reduction In System Identification
1986 (English)In: Proceedings of the American Control Conference, 1986, no New York, NY, USA, 1260-1266 p.Conference paper (Refereed)
It is shown how to use model reduction in system identification. An identification algorithm is proposed which is to be based on the least-squares identification method and either of the three model-reduction techniques: frequency-weighted L**2 model reduction, model reduction via a frequency-weighted balanced realization or frequency-weighted optimal Hankel-norm model reduction. The frequency-weighted L**2 model reduction is optimal in a minimum variance sense, while the advantage of the two other model-reduction techniques is that a consistent identification algorithm with a closed-form solution is obtained.
Place, publisher, year, edition, pages
1986. no New York, NY, USA, 1260-1266 p.
MATHEMATICAL TECHNIQUES - Least Squares Approximations, MINIMUM VARIANCE, MODEL REDUCTION, CONTROL SYSTEMS
IdentifiersURN: urn:nbn:se:kth:diva-55457OAI: oai:DiVA.org:kth-55457DiVA: diva2:471557
1986 American Control Conference. Seattle, WA, USA
QC 20120104. Sponsors: American Automatic Control Council, Research Triangle Park, NC, U; AIAA, New York, NY, USA; AIChE, New York, NY, USA; ASME, New York, NY, USA; IEEE, New York, NY, USA; et al,2012-01-022012-01-022013-09-05Bibliographically approved