H∞ - based fault diagnosis for diesel engines

Boulaïd Boulkroune (1) , Olivier Pages (1) , Abdel Aitouche (2) , Ali Zemouche (3) , Ahmed El Hajjaji (1)

1 MIS - Modélisation, Information et Systèmes - UR UPJV 4290
2 LAGIS - Laboratoire d'Automatique, Génie Informatique et Signal
3 CRAN - Centre de Recherche en Automatique de Nancy

The problem of fault detection for diesel engines is addressed in this work. The proposed approach allows to detect and estimate actuator and sensor faults. The residual design procedure is proposed for nonlinear parameter-varying systems. Inspired by some works in literature, the maximization problem of fault effect and the minimization problem of disturbances effect on the residuals is transformed to one problem of L2-norm minimization. Sufficient conditions for the existence of residual generator are formulated in terms of linear matrix inequality (LMI) sets. Performances of the proposed approach are shown through the application to a diesel engine model.

Domaines: Automatique / Robotique

Boulaïd Boulkroune, Olivier Pages, Abdel Aitouche, Ali Zemouche, Ahmed El Hajjaji. H∞ - based fault diagnosis for diesel engines. IEEE Multi-Conference on Systems and Control, MSC 2014, Oct 2014, Antibes, France. ⟨10.1109/CCA.2014.6981414⟩. ⟨hal-01059478⟩