DocumentCode :
3545309
Title :
Dynamical approach of Riccati difference equations to non linear filters stability in constrained state estimation systems
Author :
Elizabeth, S. ; Jothilakshmi, R.
Author_Institution :
Dept. of Math., Auxilium Coll., Vellore, India
fYear :
2012
fDate :
23-25 Aug. 2012
Firstpage :
141
Lastpage :
147
Abstract :
In this paper the stability of discrete time Extended Kalman Filter (EKF) when applied to non linear system with state estimation constraints are discussed. The stochastic stability of the constrained extended Kalman filter is considered then the analysis is extended to the estimation error-based constrained extended Kalman filter. The estimation error of the EKF with known constraints on the states remains bounded when the initial error and noise terms are small, and the solution of the Riccati difference equation remains positive definite and bounded. This leads to convergence of the filter and its stability. It is very sensitive to initialization and filter divergence is inevitable if the arbitrary noise matrices have not chosen appropriately.
Keywords :
Kalman filters; Riccati equations; difference equations; matrix algebra; nonlinear filters; numerical stability; state estimation; stochastic processes; Riccati difference equations; arbitrary noise matrices; constrained state estimation system; convergence; discrete time extended Kalman filter stability; estimation error-based constrained extended Kalman filter; filter divergence; initialization; nonlinear filter stability; stochastic stability; Maximum likelihood detection; Nonlinear filters; Extended Kalman Filter; Riccati Difference equations; State Estimation; Stochastic Stability;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Advanced Communication Control and Computing Technologies (ICACCCT), 2012 IEEE International Conference on
Conference_Location :
Ramanathapuram
Print_ISBN :
978-1-4673-2045-0
Type :
conf
DOI :
10.1109/ICACCCT.2012.6320758
Filename :
6320758
Link To Document :
بازگشت