DocumentCode :
3050498
Title :
Nonstationary stochastic realization and single sample identification
Author :
Benveniste, A. ; Fuchs, J.-J.
Author_Institution :
IRISA/INRIA, Rennes C??dex, France
fYear :
1983
fDate :
- Dec. 1983
Firstpage :
311
Lastpage :
316
Abstract :
Gauss-Markov processes excited by nonstationary noises are encountered in the modelling of vibrating systems. We prove that the classical Instrumental Variable method, as well as the Ho-Kalman realization algorithm, for identifying the pole part (modal characteristics) of the model, are consistent when used on a single sample of the (nonstationary) signal.
Keywords :
Econometrics; Functional analysis; Gaussian processes; Instruments; Signal analysis; Signal processing; Stochastic processes; Stochastic resonance; Stochastic systems; White noise;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Decision and Control, 1983. The 22nd IEEE Conference on
Conference_Location :
San Antonio, TX, USA
Type :
conf
DOI :
10.1109/CDC.1983.269850
Filename :
4047556
Link To Document :
بازگشت