Skip to Main content Skip to Navigation

Observateurs non linéaires pour les systèmes à mesures asynchrones : application robotique mobile

Abstract : The aim of observability studies and observer design is to reconstruct the state of a dynamic system using the measurements available. In particular, the Kalman filter algorithm is considered. This widely-studied and used observer exists in several versions : for linear or nonlinear systems, for discrete, continuous or even continuous-discrete time, in the stochastic or deterministic framework. However, Most of the time, these observers are used with the assumption that the measurements provided by the sensors are synchronous. Most of the time, this assumption can be far from the physical reality, in particular when dealing with robotic systems. In this memoir, an observer tailored for nonlinear continuous-discrete asynchronous systems is presented. These systems are made of a continuous state equation and a multirate sampled output equation. Based on the existing high-gain Extended Kalman Filter for continuous nonlinear systems and continuous-discrete nonlinear systems with synchronous outputs, we develop an ad-hoc formalism and design an observer with a deterministic point of view. Its convergence is proven analytically and illustrated by an application on a mobile robotic system.
Document type :
Complete list of metadata
Contributor : ABES STAR :  Contact
Submitted on : Friday, May 21, 2021 - 2:51:23 PM
Last modification on : Sunday, June 26, 2022 - 10:20:17 AM
Long-term archiving on: : Sunday, August 22, 2021 - 6:47:17 PM


Version validated by the jury (STAR)


  • HAL Id : tel-03232332, version 1


Aïda Feddaoui-Papin. Observateurs non linéaires pour les systèmes à mesures asynchrones : application robotique mobile. Automatique / Robotique. Université de Toulon, 2020. Français. ⟨NNT : 2020TOUL0008⟩. ⟨tel-03232332⟩



Record views


Files downloads