Feedback linearization-based MIMO model predictive control with defined pseudo-reference for hydrogen regulation of automotive fuel cells

Abstract

Proton exchange membrane fuel cells (PEMFCs) possess the benefits of high conversion efficiency, low operation noise and no pollution which are considered as potential power solutions for the automotive application. The supply and recirculation of hydrogen determines the output performance of PEMFC that is required to be well regulated. The hydrogen excess ratio (HER) is asked to locate at appropriate operating points and the pressure difference between anode and cathode should be limited within a reasonable range. In this paper, a pseudo-reference-based multiple-input multiple-output (MIMO) model predictive control (MPC) integrated with the feedback linearization is proposed for the HER regulation and the pressure balance of electrodes. The nonlinear hydrogen delivery system model is developed which consists of two actuators: flow control valve and hydrogen circulating pump. To address the high nonlinearity of the hydrogen delivery system, the feedback linearization is formulated. The MIMO MPC is then developed based on the feedback linearized model with the defined pseudo-reference which can effectively reduce the overshoot of HER. Comparing with the normal MPC based on one working-point first-order Taylor expansion, the more accurate hydrogen regulation performance of the proposed MPC can be achieved. The proposed MPC can effectively reduce the HER overshoot by 59.25% at the initial response phase with nearly same convergence time. Under the New European Driving Cycle (NEDC) simulation condition, the average overshoot reduction of anode pressure and HER can reach to 47.45% and 77.34% that demonstrates a well application prospect of the proposed MPC in the automotive PEMFC hydrogen regulation. A hardware in loop (HIL) experiment was finally conducted to show the real-time capacity of the proposed MPC method.