Skip to Main content Skip to Navigation

Estimation Temps Réel des Etats Dynamiques d'un Véhicule Automobile

Abstract : The problem is the real time estimation of dynamic vehicle states using specific sensors sets we search to determine. These states correspond to the vertical velocity on each car corner and each wheel. It also includes the relative displacement and velocity of the suspension and tire deflection. In a fist part, taking into account all the vehicle situations, we have built a specific dynamic vehicle model based on mechanical equations. In this part some simplifications and approximations have been realized to adapt the algorithm to a real time application with its Electronic Control Unit limitations. Then we have adjusted the model parameters by using vehicle data from road measurements. In the second part, we have built filtering algorithms based on different sensor sets: a first setting using only four suspension relative displacement sensors, a second one using the previous sensors with an accelerometer on each corner of the car, a third one using the four relative displacement sensors and four wheel accelerometers. This study leads to express differently the evolution of the vehicle state variables, to decouple the complete model into four sub-systems and design a specific algorithm for each of them. To conclude we have tested each algorithm in simulation from measured vehicle data on reference road. This step leads to quantify the performances of the three filters. The conclusion is that the third configuration is the best sensor set for controlled suspension system applications. The outcome of this work is the ability to construct an algorithm embedded on actual Electronic Control Units available in automotive industry. Also, this work showed that it's preferable to install accelerometers on wheel rather than on each vehicle's corner.
Document type :
Complete list of metadatas
Contributor : Josiane Dartron <>
Submitted on : Thursday, December 3, 2009 - 11:19:35 AM
Last modification on : Monday, December 14, 2020 - 12:38:04 PM
Long-term archiving on: : Tuesday, October 16, 2012 - 3:15:15 PM


  • HAL Id : tel-00438292, version 1



Sébastien Plancke. Estimation Temps Réel des Etats Dynamiques d'un Véhicule Automobile. Automatique / Robotique. Université Paris Sud - Paris XI, 2009. Français. ⟨tel-00438292⟩



Record views


Files downloads