Research On Modeling And Control Of Fuel Cell Engine System

The proton exchange membrane fuel cell (PEMFC) has many superior characteristics, such as high power density, low operating temperature, high efficiency, low noise and so on. It has received widespread attention in many countries. The development of research on modeling and control of fuel cell engine system is the important way to enhance its performance, also is one of the important directions in fuel cell research area. In view of this hot topic, on the foundation of analyzing fuel cell engine system structure, the PEMFC engine is studied by neural network model, and temperature fuzzy control system and air supplies system predicted control algorithm are proposed. The main contents are as follows:Firstly, research survey on modeling and control of PEMFC engine system is introduced, and PEMFC work principle, engine system structure and control characteristic is analyzed, then two methods are contrasted between mechanism modeling and identification modeling.In order to study the relations between each parameter of engine system and output performance, temperature, hydrogen pressure, stack current are chosen as system input, stack output voltage as system output. Fuel cell engine system input-output model is established based on the BP neural network which has approaching ability to the non-linear relations, then the basis for the formulation control strategy is provided.In view of overshoot in PEMFC temperature control, the union control plan which union electric heating, air cooling and water cooling is proposed. Then temperature control algorithm based on the fuzzy neural network is studied, which can adjust input degree of membership, and strengthen control rule robustness, reduce the temperature fluctuation, and solve the problem of slowly start from low temperature. Also, the more ideal control effect in the high temperature situation is given.In view of lag problem in the fuel cell engine wind supply, model predicted control algorithm based on the neural network is proposed, which realizes fuel cell engine air supplies optimized control, and enhances the air supplies dynamic response ability. The simulation result indicated that the control proposal may obtain better control effect.