DocumentCode :
1112731
Title :
An ARMA robust system identification using a generalized lp norm estimation algorithm
Author :
Chen, Bor-Sen ; Chen, Jeng-Ming ; Shern, Shen-Ching
Author_Institution :
Dept. of Electr. Eng., Nat. Tsing Hua Univ., Hsinchu, Taiwan
Volume :
42
Issue :
5
fYear :
1994
fDate :
5/1/1994 12:00:00 AM
Firstpage :
1063
Lastpage :
1073
Abstract :
The parameter estimation of ARMA systems with noisy output data is considered. The system output is corrupted by measurement noise, and the noise distribution is assumed to be unknown. A generalized lp norm iterative estimation algorithm (1<p<∞) is proposed to approach the maximum likelihood estimate of system parameters according to the noise distribution. Since the exponent p of lp norm estimation algorithm is sensitive to the noise distribution, based on the sample kurtosis of the residual, an adequate exponent p of lp norm estimation algorithm can be selected to achieve the efficient parameter estimation at each iteration step. Finally, several simulation results are presented to illustrate the proposed lp norm estimation algorithm; and the authors find that the proposed generalized lp norm estimation algorithm offers significant advantage to robust system parameter estimation problem with unknown noise distribution
Keywords :
iterative methods; linear systems; maximum likelihood estimation; noise; parameter estimation; stochastic processes; time series; ARMA systems; generalized lp norm estimation algorithm; iterative estimation algorithm; maximum likelihood estimate; measurement nois; noise distribution; noisy output data; parameter estimation; robust system identification; simulation results; system output; system parameters; Gaussian distribution; Gaussian noise; Iterative algorithms; Least squares approximation; Maximum likelihood estimation; Noise measurement; Noise robustness; Parameter estimation; Signal processing algorithms; System identification;
fLanguage :
English
Journal_Title :
Signal Processing, IEEE Transactions on
Publisher :
ieee
ISSN :
1053-587X
Type :
jour
DOI :
10.1109/78.295211
Filename :
295211
Link To Document :
بازگشت