Polymer electrolyte membrane fuel cell fault diagnosis based on empirical mode decomposition

Cédric Damour (1) , Michel Benne (1) , Brigitte Grondin-Perez (1) , Miloud Bessafi (1) , Daniel Hissel (2) , Jean-Pierre Chabriat (1)

1 ENERGY Lab - Energy Lab
2 FC LAB - Laboratoire Commun de Belfort : Hydrogène et Pile à Combustible pour les applications au transport

Diagnosis tool for water management is relevant to improve the reliability and lifetime of polymer electrolyte membrane fuel cells (PEMFCs). This paper presents a novel signal-based diagnosis approach, based on Empirical Mode Decomposition (EMD), dedicated to PEMFCs. EMD is an empirical, intuitive, direct and adaptive signal processing method, without pre-determined basis functions. The proposed diagnosis approach relies on the decomposition of FC output voltage to detect and isolate flooding and drying faults. The low computational cost of EMD, the reduced number of required measurements, and the high diagnosis accuracy of flooding and drying faults diagnosis make this approach a promising online diagnosis tool for PEMFC degraded modes management.

Mots clés: Empirical mode decomposition; Flooding; Drying; PEM fuel cell; Non-invasive fault diagnosis

Domaines: Génie des procédés

Cédric Damour, Michel Benne, Brigitte Grondin-Perez, Miloud Bessafi, Daniel Hissel, et al.. Polymer electrolyte membrane fuel cell fault diagnosis based on empirical mode decomposition. Journal of Power Sources, 2015, 299, pp.596--603. ⟨10.1016/j.jpowsour.2015.09.041⟩. ⟨hal-01221973⟩