Comparison of observer approaches for actuator fault estimation in semi-active suspension systems - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

Comparison of observer approaches for actuator fault estimation in semi-active suspension systems

Manh Quan Nguyen
  • Fonction : Auteur
  • PersonId : 959949
  • IdRef : 168065703
Olivier Sename
Luc Dugard

Résumé

In this paper, the actuator fault estimation problem of semi-active suspension systems is considered. For instance, an oil leakage in the damper could cause a reduction of the damping force. The fault estimation requires a modeling of the damper fault (both multiplicative and additive fault models can be used). Three observer-based approaches are compared for fault estimation: an observer using fast adaptive fault estimation (FAFE) approach (used for estimation of additive faults), a parametric adaptive observer (AO) and a switched LPV observer (LPVO) (both intended to estimate mulplicative faults); Since the damper fault estimation is strongly affected by the unknown road disturbances, an H ∞ performance objective is used to reduce the effect of disturbances on the estimation error for performance assessment. Some simulations are performed on a quarter car model to validate these methodologies and a comparison is then given to shows the interest of each method. Keywords: Fault estimation, semi-active damper fault, adaptive observer, fast fault adaptive estimation, LPV observer .
Fichier principal
Vignette du fichier
systol2016 provisoire.pdf (393.74 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01361742 , version 1 (07-09-2016)

Identifiants

Citer

Manh Quan Nguyen, Olivier Sename, Luc Dugard. Comparison of observer approaches for actuator fault estimation in semi-active suspension systems. SysTol 2016 - 3rd International Conference on Control and Fault-Tolerant Systems, Sep 2016, Barcelone, Spain. ⟨10.1109/SYSTOL.2016.7739755⟩. ⟨hal-01361742⟩
258 Consultations
175 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More