Title :
ARMA system identification: An algebraic approach
Author :
Cadzow, James A. ; Soloman, O.
Author_Institution :
Arizona State University, Tempe, Arizona, USA
Abstract :
An algebraic procedure for linear system identification using empirically obtained noise contaminated, excitation - response data is proposed. This method entails an eigen analysis of a combined excitation- response correlation matrix estimate. This analysis provides both model order information as well as the required model parameter estimates.
Keywords :
Additive noise; Autocorrelation; Contracts; Information analysis; Least squares methods; Linear systems; Nonlinear equations; Parameter estimation; System identification; Vectors;
Conference_Titel :
Acoustics, Speech, and Signal Processing, IEEE International Conference on ICASSP '84.
DOI :
10.1109/ICASSP.1984.1172706
