The Effects of Rapid Sampling in System Identification
1990 (English)In: Automatica, ISSN 0005-1098, Vol. 26, no 1, 167-170 p.Article in journal (Refereed) Published
This paper deals with the effects of rapid sampling in system identification. In continuous time, noise models of non-zero relative degree imply that one has to differentiate the measured data to find the prediction error parameter estimate. This problem can be avoided by using noise models of relative degree zero or by introducing prefilters. The corresponding difficulty for the discrete time case is less obvious. However, by linking the discrete time and continuous time parameter estimation problem, we show that the same problem arises for rapid sampled systems.
Place, publisher, year, edition, pages
Elsevier, 1990. Vol. 26, no 1, 167-170 p.
System identification, Parameter estimation, Sampled data systems, Stochastic systems, Statistics, (Rapid sampling)
IdentifiersURN: urn:nbn:se:liu:diva-100538DOI: 10.1016/0005-1098(90)90167-GOAI: oai:DiVA.org:liu-100538DiVA: diva2:663053