Online implementation of SVM based fault diagnosis strategy for PEMFC system

Zhongliang Li (1, 2, 3) , Rachid Outbib (3) , Stefan Giurgea (1, 2) , Daniel Hissel (1, 2) , Samir Jemei (2)

1 FC LAB - Laboratoire Commun de Belfort : Hydrogène et Pile à Combustible pour les applications au transport
2 FEMTO-ST - Franche-Comté Électronique Mécanique, Thermique et Optique - Sciences et Technologies (UMR 6174)
3 LSIS - Laboratoire des Sciences de l'Information et des Systèmes

This paper deals with the online diagnosis of Polymer Electrolyte Membrane Fuel Cell (PEMFC) systems. The pattern classification tool Support Vector Machine (SVM) is used to achieve fault detection and isolation (FDI). The algorithm is integrated into an embedded system of the type System in Package (SiP) and validated online in an experimental platform. Four concerned faults are diagnosed successfully online. Additionally, a procedure is proposed to improve the performance of robustness and raise the diagnosis accuracy.

Mots clés: Robustness; SVM; Fault diagnosis; Online implementation; PEMFC systems

Domaines: Automatique / Robotique; Energie électrique

Zhongliang Li, Rachid Outbib, Stefan Giurgea, Daniel Hissel, Samir Jemei. Online implementation of SVM based fault diagnosis strategy for PEMFC system. the 6th International Conference on ”Fundamentals & Development of Fuel Cells” (FDFC), Feb 2015, Toulouse, France. ⟨hal-01113589⟩