Assessing Model Quality from Data
1991 (English)Report (Other academic)
The problem of deriving error bounds for estimated transfer functions is addressed. By blending a priori knowledge and information obtained from measured data, we show how the uncertainty of transfer function estimates can be quantified. The emphasis is on errors due to model mismatch. The effects of unmodeled dynamics can be considered as bounded disturbances. Hence, techniques from set membership identification can be applied to this problem. The approach taken corresponds to weighted least squares estimation, and provides hard frequency domain transfer function error bounds.
Real processes rarely are time-invariant. Hence, the unmodeled dynamics contains a time-varying part. It is important to quantify this model error as well. Herein, this is done in terms of confidence bounds for the “frozen” transfer function, i.e. the sequence of transfer functions obtained when freezing the time variable at succesive times. This method is based on the assumption that the true system is varying around some nominal system.
Place, publisher, year, edition, pages
Linköping: Linköping University , 1991.
LiTH-ISY-I, ISSN 8765-4321 ; 1187
Error bounds, Transfer functions, Model mismatch, Quality
IdentifiersURN: urn:nbn:se:liu:diva-55346OAI: oai:DiVA.org:liu-55346DiVA: diva2:316036