Real time modelling of the dynamic mechanical behaviour of PEMFC thanks to neural networks

Modelling complex dynamic mechanical systems, such as PEMFC, without any physical models is a difficult challenge but it could allow the monitoring of endurance tests of fuel cell systems. Neural networks are recognised as powerful numerical tools for predicting complex and nonlinear dynamic behaviours. They require only data limited to experimental inputs and outputs but the choice of an adapted architecture is critical. This paper presents a method for defining a neural network architecture optimised for the fuel cell systems. The associated experimental conditions specifying the vibration tests to train and validate were defined. They consist of swept sinus as well as random excitation forces. The resulting simulations are presented and analysed.

► The interest of a neural network model adapted to the monitoring of fuel cells under vibrating conditions.
► A strategy describing step by step the methodology used to define the optimal neural architecture.
► The simulation and experimental results show the importance of the optimal neural network architecture.