Modelling and controlling of polymer electrolyte fuel cell systems

This thesis focuses on the modelling and controlling of polymer electrolyte fuelcell (PEFC) systems. A system level dynamic PEFC model has been developedto test the system performance (output voltage, reactants gas partial pressures,and stack temperature) for different operating conditions. The simulation resultsare in good agreement with the experimental data, which indicates that thePEFC model is well qualified to capture the dynamic performance of the PEFCsystem. Controlling strategies play a significant role in improving the fuel cellsystem’s reliability. Novel model predictive control (MPC) controllers and proportional–integral–derivative (PID) controllers are proposed and implemented indifferent PEFC systems to control voltage and regulate temperature to enhancesystem performance. MPC controllers show superior performance to PID controllers in tracking the reference value, with less overshoot and faster response. Anovel hydrogen selective membrane reactor (MR) is designed for methanol steamreforming (MSR) to produce fuel cell grade hydrogen for PEFC stack use. Thebackpropagation (BP) neural network algorithm is applied to find the mappingrelation between the MR’s operating parameters and the PEFC system’s outputperformance. Simulation results show that the BP neural network algorithm canwell predict the system behaviour and that the developed mapping relation modelcan be used for practical operation guidance and future control applications