Fuel cell fault diagnosis: A stochastic approach

This paper deals with the problem of fault diagnosis for fuel cells. More precisely, the goal is to present a new strategy based on a stochastic approach to fault diagnosis which gives additional information of fuel cell conditions and operation modes, making fault diagnosis procedure simpler to achieve by using cell voltage probability density functions as clustering parameter. Flooding, and electrode contamination failure classification are discussed