Robust $H_\infty$ proportional-integral observer for fault diagnosis: Application to vehicle suspension

Manh-Hung Do (1) , Damien Koenig (1) , Didier Theilliol (2)

1 GIPSA-SLR - GIPSA - Systèmes linéaires et robustesse
2 CRAN - Centre de Recherche en Automatique de Nancy

The main contribution of this paper is the design of robust $\mathcal{H}_{\infty}$ proportional-integral (PI) observer for fault detection, isolation and magnitude estimation, applied for uncertain linear time-invariant (LTI) system. It is based on a combination of PI observer and $\mathcal{H}_{\infty}$ norm in order to deal with system uncertainty and disturbance attenuation. This multi-objective problem is tackled by the solution of linear matrix inequality (LMI) feasibility problem, obtained from the analysis of estimation errors and the use of majoration lemma. The application to a real suspension system is then presented to highlight the performances of the proposed observer.

Mots clés: Robust estimation.; Fault diagnosis; Observers; Uncertain linear systems; Fault detection; Semi-active suspension

Domaines: Automatique / Robotique

Manh-Hung Do, Damien Koenig, Didier Theilliol. Robust $H_\infty$ proportional-integral observer for fault diagnosis: Application to vehicle suspension. 10th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes, SAFEPROCESS 2018, Aug 2018, Varsovie, Poland. ⟨hal-01830165⟩