Title :
Sensitivity analysis of linear optimal stochastic observers
Author :
Madjarov, Nikola ; Mihaylova, Ludmila
Author_Institution :
Dept. of Autom., Tech. Univ. Sofia, Bulgaria
Abstract :
The paper considers the continuous-time and discrete-time Kalman filters under noise uncertainties. The influence of these uncertainties on the filter performance can be characterized by the so-called sensitivity functions. Different ways for defining sensitivity functions are proposed and relationships for the corresponding functions are derived with the aid of matrix derivatives and finite differences. The results are illustrated with examples
Keywords :
Kalman filters; Riccati equations; continuous time filters; discrete time filters; observers; sensitivity analysis; continuous-time Kalman filters; discrete-time Kalman filters; finite differences; linear optimal stochastic observers; matrix derivatives; noise uncertainties; sensitivity analysis; sensitivity functions; Covariance matrix; Estimation error; Filtering algorithms; Kalman filters; Mathematical model; Riccati equations; Sensitivity analysis; State estimation; Stochastic processes; Uncertain systems;
Conference_Titel :
Systems, Man and Cybernetics, 1993. 'Systems Engineering in the Service of Humans', Conference Proceedings., International Conference on
Conference_Location :
Le Touquet
Print_ISBN :
0-7803-0911-1
DOI :
10.1109/ICSMC.1993.384919
