DocumentCode :
835534
Title :
Multidimensional maximum likelihood failure detection and estimation
Author :
Friedland, Bernard
Author_Institution :
Singer Company, Little Falls, NJ, USA
Volume :
26
Issue :
2
fYear :
1981
fDate :
4/1/1981 12:00:00 AM
Firstpage :
567
Lastpage :
570
Abstract :
An algorithm for maximum likelihood failure detection and estimation is provided for multidimensional processes as an extension of the theory given in [1]. It is shown that failure identification is enhanced by selecting the observation matrix 
to satisfy an orthogonality condition.
to satisfy an orthogonality condition.Keywords :
Fault diagnosis; Maximum-likelihood detection; maximum-likelihood (ML) estimation; Inertial navigation; Maximum likelihood detection; Maximum likelihood estimation; Multidimensional systems; Noise measurement; Random variables; Smoothing methods; State estimation; System performance; System testing;
fLanguage :
English
Journal_Title :
Automatic Control, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9286
Type :
jour
DOI :
10.1109/TAC.1981.1102638
Filename :
1102638
Link To Document :
