Model predictive control of water management in PEMFC

Abstract

Water management is critical for Proton Exchange Membrane Fuel Cells (PEMFC). An appropriate humidity condition not only can improve the performances and efficiency of the fuel cell, but can also prevent irreversible degradation of internal composition such as the catalyst or the membrane. In this paper we built the model of water management systems which consist of stack voltage model, water balance equation in anode and cathode, and water transport process in membrane. Based on this model, model predictive control mechanism was proposed by utilizing Recurrent Neural Network (RNN) optimization. The models and model predictive controller have been implemented in the MATLAB and SIMULINK environment. Simulation results showed that this approach can avoid fluctuation of water concentration in cathode and can extend the lifetime of PEM fuel cell stack.