Analysis and Diagnosis of Faults in the PEMFC for Fuel cell Electrical Vehicles

In recent years, according to the upcoming challenge of pollution, fuel saving, to use on FCEV is increasing. It can be that fuel cell power train divided in the PEMFC, Batteries, DC/DC converters, DC/AC inverters and electrical motors. The Proton Exchange Membrane Fuel cells (PEMFC) have consistently been considered for transportation application. Characteristic features of PEMFC include lower temperature (50 to 100 deg. C) and solid polymer electrolyte membrane. In this work, experiments have shown that the temperature distributions can significant influence on the performance of the PEMFC. Also analytical studies have indicated improvement of ionic resistivity of the electrolyte membrane, kinetics of electrochemical reaction and gas diffusion electrodes have directly related to temperature. This work evaluated the effectiveness of temperature on a single and stack fuel cell. In addition, a 3D model is developed by effective of temperature on performance on the fuel cell. In this thesis, two PEM fuel cells have been considered to find out the relationship and analyze the behaviors of the cell voltage and temperature distributions under various operating conditions. An experimental study for voltage and temperature has been executed, using one cell, 12 thermocouples and 12 voltage sensors have been installed at different points of the cell. In this work a new model was proposed to improve the lifetime and reliability of the power train and to detect online faults. Besides, current distributions in different points of the cell based on varying operating conditions are calculated by the Newton Raphson method. On the basis of the developed fault sensitive models above, an ANN based fault detection; diagnosis strategy and the related algorithm have been developed. The identified patterns ANN have been used in the supervision and the diagnosis of the PEMFC drive train. The ANN advantages of the ability to include a lot of data made possible to classify the faults in terms of their type. (author)