A Novel Adaptive Model Predictive Control for Proton Exchange Membrane Fuel Cell in DC Microgrids

In this paper, a novel adaptive model predictive control (AMPC) algorithm is proposed for proton exchange membrane fuel cells (PEMFCs) to improve its output power tracking performance in DC microgrids. The proposed AMPC algorithm is capable of producing superior PEMFC control performance over the conventional MPC and PI controllers. Compared with PI control, AMPC is able to handle physical constraints, which ensures the safe operation of PEMFC under different conditions. Furthermore, it overcomes the common problem of system model dependence that is shared by nearly all the model-based control methods. The convergence of parameter estimation in the proposed AMPC is rigorously proved. The effectiveness of the proposed algorithm under various operating conditions and system parameter variations are verified by simulation results.