Evaluation of estimation quality with respect to sensors losses

Vincent Sircoulomb (1, 2) , Ghaleb Hoblos (2) , Houcine Chafouk (2) , José Ragot (1)

1 CRAN - Centre de Recherche en Automatique de Nancy
2 IRSEEM - Institut de Recherche en Systèmes Electroniques Embarqués

This paper deals with the estimation quality of nonlinear filters in relation to sensors losses. In a nonlinear state space representation setting, Central Difference Kalman Filter, Ensemble Kaman Filter and Particle Filter are tested on a second order system. Their comportment in relation to the sensors available are then studied, in order to compare the estimation quality by computing criteria such that variances of estimation errors and observability gramian.

Mots clés: Nonlinear; Kalman Filter; Monte Carlo filter; Observability; Gramian

Domaines: Automatique / Robotique

Vincent Sircoulomb, Ghaleb Hoblos, Houcine Chafouk, José Ragot. Evaluation of estimation quality with respect to sensors losses. 7ème Congrès International Pluridisciplinaire Qualité et Sûreté de Fonctionnement, Qualita 2007, Mar 2007, Tanger, Morocco. pp.CDROM. ⟨hal-00200863⟩