A novel cascade control of PEMFC: Regulation of stack voltage and air breathing subsystem

The proton exchange membrane fuel cell (PEMFC) stack is an excellent source of clean energy for electric vehicles with zero carbon footprint. The abrupt changes in power demand in automotive applications can adversely impact the oxygen excess ratio (OER) and stack voltage. Therefore, this study aims to design an intelligent cascade control framework to ensure the durable and efficient operation of PEMFC. A novel mixed locally recurrent fuzzy neural network proportional-integral-derivative (RFNN-PID) control architecture is proposed. The RFNN-PID and proportional-integral (PI) controllers are employed as the primary and secondary controllers in cascaded loops, respectively. Four other cascaded schemes with fuzzy PID and PI (FPID/PI), two-degree-of-freedom PID & two-degree-of-freedom PI (2DOF-PID/2DOF-PI), fractional order PID & fractional order PI (FOPID/FOPI), and PID/PI controllers are designed for comparison. The controller parameters are optimised using the mayfly algorithm (MA) to minimise the integral time absolute error (ITAE). The findings suggest the cascaded RFNN-PID/PI strategy yields the fastest voltage regulation and minimum ITAE (0.3018 s & 3.406), as compared to cascaded FPID/PI (0.4087 s & 4.273), 2DOF-PID/2DOF-PI (0.4213 s & 6.626), FOPID/FOPI (0.8135 s & 9.083) and PID/PI (1.066 s & 10.89). The proposed controller also exhibits robust performance in the presence of disturbance and noise. Hence, it is concluded that the adaptive RFNN-PID/PI controller efficiently controls the voltage and regulates OER even in adverse operating conditions.