Optimized Operation of Fuel Cell Systems Based on a Two-Layer Control Structure

With the increasing demand for environmentally friendly and efficient energy conversion methods, proton exchange membrane fuel cells (PEMFCs) have become a great choice, because they are efficient, adaptable and environmentally friendly. However, the complex relationship between fuel cell stack and its auxiliary components, such as air compressor, humidifier and cooling system, has brought great problems: it is necessary to ensure rapid response and control costs. Finally, we established a complete PEMFC system model in MATLAB/Simulink, which is completely nonlinear and based on physical principles. It can capture the gas flow of anode and cathode, the water transmission through proton exchange membrane, the thermal change of the stack and the operation of auxiliary equipment. Using this model, we propose a two-level economic predictive control scheme, which combines real-time optimization (RTO) and model predictive control (MPC). The upper RTO uses particle swarm optimization (PSO) algorithm to maximize the net power output of the system by calculating the optimal set points of stack voltage, cathode humidity level and operating temperature. The lower MPC ensures that these set points can be followed quickly and accurately even if the load conditions change. The simulation results show that the RTO-MPC method proposed by us significantly improves the dynamic performance and operation economy of PEMFC system. Especially when the current demand changes suddenly, this method shortens the time required for stabilization, reduces the transient overshoot and reduces the integral absolute error (IAE).