Operating a proton exchange membrane fuel cell (PEMFC) system to produce power at the maximum power efficiency is one of the key issues in PEMFC's wide-spread applications. However, power density exhibits complex behavior and nonlinear dynamics with respect to the output cell voltage, fuel cell temperature, anode and cathode pressure, inlet gas humidity, and so on. In this paper, the distribution of power density in the domain of the output cell voltage and fuel cell temperature is delineated. By this delineation, the quadratic polynomial fitting was used to approximate the power density curve in local interval and estimate the maximum power efficiency point. Generalized predictive control (GPC) is presented to overcome the problem of time-varying dynamics of PEMFC in real time via applying a forgetting factor recursive least square (FFRLS) method. Based on the approximation and generalized predictive control strategy, maximum power efficiency operation of PEMFC is applied. The results of this work can contribute to the operation of PEMFC at the maximum power point, which guarantees the plant generating maximum power at the lowest consumption of hydrogen.
Read full abstract