Improving dynamic performance of proton-exchange membrane fuel cell system using time delay control

Transient behaviour is a key parameter for the vehicular application of proton-exchange membrane (REM) fuel cell. The goal of this presentation is to construct better control technology to increase the dynamic performance of a REM fuel cell. The REM fuel cell model comprises a compressor, an injection pump, a humidifier, a cooler, inlet and outlet manifolds, and a membrane-electrode assembly. The model includes the dynamic states of current, voltage, relative humidity, stoichiometry of air and hydrogen, cathode and anode pressures, cathode and anode mass flow rates, and power. Anode recirculation is also included with the injection pump, as well as anode purging, for preventing anode flooding. A steady-state, isothermal analytical fuel cell model is constructed to analyze the mass transfer and water transportation in the membrane. In order to prevent the starvation of air and flooding in a REM fuel cell, time delay control is suggested to regulate the optimum stoichiometry of oxygen and hydrogen, even when there are dynamical fluctuations of the required REM fuel cell power. To prove the dynamical performance improvement of the present method, feed-forward control and Linear Quadratic Gaussian (LQG) control with a state estimator are compared. Matlab/Simulink simulation is performed to validate the proposed methodology to increase the dynamic performance of a PEM fuel cell system. (C) 2010 Elsevier B.V. All rights reserved.