Model Reduction of Estimated Models
1988 (English)Report (Other academic)
This paper deals with the connection between system identification and model reduction. We will from a statistical point of view discuss how to reduce the order of high-order models obtained from an identification experiment. We will apply these results to estimate transfer functions by means of a high-order FIR model and model reduction. The model reduction techniques considered are: Frequency weighted L2-norm model reduction and model reduction via a truncated frequency weighted balanced realization.
Place, publisher, year, edition, pages
Linköping: Linköping University , 1988. , 6 p.
LiTH-ISY-I, ISSN 8765-4321 ; 906
Mathematical statistics, Monte Carlo methods, Approximation theory, Maximum likelihood, Control systems
IdentifiersURN: urn:nbn:se:liu:diva-104209OAI: oai:DiVA.org:liu-104209DiVA: diva2:695361