An unknown input extended Kalman filter for nonlinear stochastic systems

Luc Meyer (1) , Dalil Ichalal (2) , Vincent Vigneron (2)

1 DTIS, ONERA, Université Paris Saclay [Palaiseau]
2 IBISC - Informatique, BioInformatique, Systèmes Complexes

This paper proposes an Unknown Input Extended Kalman Filter (UIEKF) for stochastic non linear systems affected by Gaussian noises and Unknown Inputs (UI) in both state and measurement equations. The proposed approach is based on a total decoupling of the UI, in spite of the presence of nonlinearities in the measurement equation. The UI is decoupled under some structural constraints, and a state estimator is provided. Besides an UI estimator is also proposed. Finally, the proposed filter is applied on a classical navigation example, illustrating its advantages.

Mots clés: Unknown inputs; Extended Kalman filter; State estimation; Aerospace navigation

Domaines: Automatique / Robotique; Traitement du signal et de l'image [eess.SP]

Luc Meyer, Dalil Ichalal, Vincent Vigneron. An unknown input extended Kalman filter for nonlinear stochastic systems. European Journal of Control, 2020, 56, pp.51--61. ⟨10.1016/j.ejcon.2020.01.009⟩. ⟨hal-02513621⟩