DocumentCode :
1040636
Title :
On statistical properties of a test for model structure selection using the extended instrumental variable approach
Author :
Nghia Duong, Hoai ; Landau, Ioan Dore
Author_Institution :
Lab. d´´Autom., CNRS, Grenoble, France
Volume :
39
Issue :
1
fYear :
1994
fDate :
1/1/1994 12:00:00 AM
Firstpage :
211
Lastpage :
215
Abstract :
A procedure for structural estimation using the extended instrumental variable technique is presented. The statistical distribution of the test quantity is obtained in the white noise case. A criterion is defined by introducing a penalty term on the decreasing rate (with the number of parameters) of the proposed test quantity. The consistency of the estimate is proven. Monte Carlo simulations are included
Keywords :
identification; statistical analysis; white noise; Monte Carlo simulations; extended instrumental variable approach; model structure selection test; penalty term; statistical distribution; statistical properties; structural estimation; white noise; Automatic testing; Computational efficiency; Instruments; Least squares methods; MIMO; Parameter estimation; Predictive models; Statistical distributions; System testing; White noise;
fLanguage :
English
Journal_Title :
Automatic Control, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9286
Type :
jour
DOI :
10.1109/9.273371
Filename :
273371
Link To Document :
بازگشت