Design and evaluation of adaptive neural fuzzy-based pressure control for PEM fuel cell system

Abstract

In the proton-exchange membrane fuel cell (PEMFC), the partial pressure equalization between hydrogen and oxygen is one of the most significant problems influencing PEMFC efficiency and lifetime. In this paper, an adaptive neural fuzzy inference system (ANFIS)-based pressure tracking controller for a PEMFC system is investigated under the variation of load current. The suggested controller is designed based on the dynamic model to enhance the efficiency and prolong the stack life of the PEMFC system. First, the pressure control system model is established on the premise of the electrochemical and react flow model. Next, the ANFIS controller is developed to further improve the transient performance and minimize the partial pressures of hydrogen and oxygen. Finally, the simulation and experimental results are provided to demonstrate the effectiveness of the suggested method. Compared to the super-twisting sliding mode controller (STSMC), the proposed ANFIS controller has increased the tracking performance by 89.5% in simulation and 72.06% in the experiment under the multi-step load current.

Keywords

• Proton-exchange membrane fuel cell
• Adaptive neural fuzzy control
• Nonlinear modeling
• Pressure equalization