Study On Control Of Proton Exchange Membrane Fuel Cell

Proton Exchange Membrane Fuel Cell (PEMFC) is a new type of fuel cell which is developed in recent years. It has high efficient energy conversion, high reliability, no pollution, lower noisy, fast startup, high power and many other advantages which make it nowadays the most widely studied and fastest developed in fuel cell research. PEMFC provide a way to solve the energy crisis and environmental pollution problems which our country confronted.The gradually maturation and commercialization of PEMFC put forward better performance for its reliability and high efficiency. The focus of this paper is the research of the control system whose performance is very important. The influence factors of PEMFC output characteristics and its control characteristics are analyzed. Then the temperature control of dew point humidifier and PEMFC stack is studied.In the design of the temperature control system of the dew point humidifier, a sort of PID controller based on neural network is put forward to solve the control problems caused by the nonlinear and large range of disturbances and two kinds of neural network for controller design is introduced in detail:BP and RBF, including their structure and training algorithm. Then the date which we got in the temperature control experiments using different controller are analyzed and comparedIn the design of the temperature control system of the PEMFC stack, the fuzzy neural network controller which has a low degree of demand for the accurate mathematical model and a good control quality is used in the paper for the difficulty of developing a mathematical model of the system. The fuzzy neural network controller which is a compound intelligence controller combining the fuzzy logical and the neural network could use its study ability to optimize logical rules in order to realize the effective control of the PEMFC stack temperature. Through the result of simulation experiment using the model set by the matlab/simulink software, fuzzy neural network controller has been proved to have a better curve in a the large range variety of systemic parameter and have good control quality and robustness which meet the demand of the PEMFC stack temperature control system.