Data-driven modeling and control of proton exchange membrane fuel cell with reducing overshoot

Proton exchange membrane fuel cell (PEMFC) is widely used in many applications such as vehicle power due to its low emission and high efficiency. Dynamic modeling and control are important for PEMFC, especially for vehicles with varying loads. In this paper, a dynamic model of PEMFC and a output voltage controller are proposed to meet the varying load. Set the inlet air temperature, inlet air pressure, inlet hydrogen pressure, cooling water temperature and cooling water pressure as the input of ANN, and the output voltage as the output of ANN which is trained by experimental data. Integral separation PID controller is used to solve the problem of excessive overshoot based on ANN model. Simulation results show that ANN can lit PEMFC system well and combine with integral separation PID controller successfully to control the output voltage of fuel cell reducing the maximum deviation by 50% compared to traditional PID controller.