Title :
Obtaining accurate confidence regions for the estimated zeros and poles in system identification problems
Author :
Vuerinckx, R. ; Pintelon, R. ; Schoukens, J. ; Rolain, Y.
Author_Institution :
Dept. ELEC, Vrije Univ., Brussels, Belgium
fDate :
4/1/2001 12:00:00 AM
Abstract :
In this note, a method is proposed to generate improved uncertainty bounds on the poles and zeros of an identified system. The uncertainty ellipsoids are replaced by dedicated uncertainty regions with a given probability level. The method can be applied to Markov estimators in the time or frequency domain
Keywords :
Markov processes; frequency-domain analysis; identification; poles and zeros; probability; time-domain analysis; uncertain systems; Markov estimators; confidence regions; dedicated uncertainty regions; estimated poles; estimated zeros; frequency domain; probability level; system identification problems; time domain; uncertainty ellipsoids; Covariance matrix; Ellipsoids; Noise generators; Noise measurement; Parameter estimation; Poles and zeros; Signal to noise ratio; System identification; Transfer functions; Uncertainty;
Journal_Title :
Automatic Control, IEEE Transactions on