MPC CONTROL OF DYNAMIC PEM FUEL CELL SYSTEM

This paper presents, an open cathode fuel cell dynamic model and an advanced control system design that allows to study the transient response of the system and its nonlinearities. The model can be used to design control strategies and design test benches to test fuel cells stacks. A dynamic performance characterization is performed (transient response to disturbances or setpoint changes), and the information is used to propose the structure of a control system capable of incorporating expert information and transition in non-linear regions. The model used in the simulation takes account of three different phenomena: (a) mass transfer, (b) thermodynamics and (c) electrochemical. The fuel cell uses hydrogen and air to generate electricity. The hydrogen is supplied to the fuel cell anode and the air is supplied to the fuel cell cathode. This model allows the calculation of the pressure and mass flow circulating in the cell. Balances are presented for cathode, anode and membrane. The energy balance proposed includes the electrochemical reaction that is exothermic and' the heat transfer to the air flowing through the cathode which has to supply the electrochemical reaction with the oxygen required and regulate the stack temperature. The ideal electric potential of a fuel cell is defined by the Nernst equation. The Nernst equation alone is not enough to describe the electrical potential during the actual operation of the system. The main sources of voltage loss are presented: Activation, electrical resistance and concentration. PEM control system is designed and simulated. A model based algorithm is used to achieve different objectives than could lead to an improvement in the system performance. Results demonstrate that the predictions in voltage and current in the fuel cell stack are close to experimental data obtained from manufacturer. Constraints in the algorithm allow avoiding hazardous operational conditions and a decrease in controlled variable variance.