A neural network model of a fuel cell stack under road vibrations

A model is very important for studying the dynamic response of the fuel cell stack under road vibrations. Mechanism models have many parameters which can´t be measured in real stack tests when the outside forces and the inside parameters are all needed to be known. The fuel cell stack used in new energy car which takes place of conventional engine is a complicated nonlinear mechanical system. while its durability under vibration is tested, only the driving and responding signals are collected. In this case, a neural network is used to predict the response of the stack under different vibration conditions and also to be used as a fault diagnose tool.