Kalman filtering utilizing future dynamics for descriptor systems

Author

Yu, Tie-Jun ; Lin, Ching-Fang ; Müller, Peter C.

Author_Institution

American GNC Corp., Chatsworth, CA, USA

Volume

1

fYear

1995

fDate

21-23 Jun 1995

Firstpage

119

Abstract

This paper studies the filtering problem of descriptor systems. The noncausal behaviour of descriptor systems leads to filtering that takes into account not only past and present dynamics, but also the future dynamics. Using the maximum likelihood estimation technique, a recursive filter for general time-varying descriptor systems is developed which makes use of past, present as well as one-step future dynamics. The existence condition of the filter is also given which is weaker than that of the filter in Nikoukhah et al. (1992) and is identical to the infinity observability in the time-invariant case

Keywords

Kalman filters; maximum likelihood estimation; observability; recursive filters; time-varying systems; Kalman filtering; descriptor systems; existence condition; general time-varying descriptor systems; infinity observability; maximum likelihood estimation technique; noncausal behaviour; one-step future dynamics; past dynamics; present dynamics; recursive filter; Covariance matrix; Estimation error; Filtering; Kalman filters; Lagrangian functions; Maximum likelihood estimation; Recursive estimation; Sufficient conditions;

fLanguage

English

Publisher

ieee

Conference_Titel

American Control Conference, Proceedings of the 1995

Conference_Location

Seattle, WA

Print_ISBN

0-7803-2445-5

Type

conf

DOI

10.1109/ACC.1995.529220

Filename

529220