Robust control of internal states in a polymer electrolyte membrane fuel cell air-feed system by considering actuator properties

Air stoichiometry, pressure, and relative humidity in the air-feed system of a vehicular polymer electrolyte membrane fuel cell (PEMFC) influence efficiency, durability and reliability. It is critical to develop robust control algorithms for these internal states to improve system performance. There is limited extant research on designing robust control algorithms that consider the three internal states as well as the constraints of real actuators, such as an air compressor, a membrane humidifier, and a back-up pressure valve (BPV). This study examines robust control strategies for the three internal states based on adaptive second order sliding mode (ASOSM) and nonlinear proportional-integral (NPI) feedback control algorithms. In the study, control targets are established based on stable properties of the PEMFC system. The study involves proposing and comparing five control strategies that are a combination of NPI and ASOSM algorithms. The following results are obtained: (1) the stable control targets for the three internal states are followed adequately by using an NPI or an ASOSM algorithm and differences only exists in dynamic processes; (2) with respect to the control of air stoichiometry, an NPI algorithm performs better than an ASOSM algorithm as chattering in air stoichiometry can be avoided and the convergence time to the target value is acceptable; (3) with respect to the control of cathodic pressure, an ASOSM algorithm performs better than an NPI algorithm as the overshoots in cathodic pressures can be effectively reduced; (4) with respect to the control of relative humidity, both NPI and ASOSM algorithms lead to a practical bang-bang strategy. The strategy that performs the best among the five strategies is selected, and the robustness of the selected strategy with respect to parameter uncertainties is verified.