A stochastic variance reduction gradient-based GSO-ANFIS optimized method for maximum power extraction of proton exchange membrane fuel cell

Proton Exchange Membrane Fuel Cells (PEMFCs) play a pivotal role in the clean energy landscape, yet their efficiency is contingent upon effective power optimization. This paper presents Maximum PowerPoint Tracking (MPPT) control schemes for PEMFCs, focusing on a ground-breaking methodology. Traditional MPPT controllers are instrumental in maintaining optimal performance; however, they often struggle with dynamic operating conditions. In response to this challenge, this research work presents a pioneering MPPT control scheme employing a stochastic variance reduction gradient system. The novelty of this approach lies in its fusion with the Glow Swarm Optimization (GSO) and the Adaptive Neuro Fuzzy Inference System (ANFIS), resulting in a robust hybrid controller. In the pursuit of optimizing the PEMFC system, the proposed GSO-ANFIS controller is subjected to rigorous testing under dynamic variations in both PEMFC temperature and load. Notably, PEMFCs, due to fluctuations in pressure and temperature, exhibit stochastic behaviour, forming a Gaussian surface. In this research, the popular Perturb and Observe (P&O) and Incremental Conductance methods are evaluated alongside the newly introduced GSO-ANFIS model. The proposed GSO-ANFIS controller outperforms its counterparts, showcasing an impressive accuracy level of 89.97%. In contrast, the Artificial Neural Network (ANN) achieves 80.33% accuracy, and the standalone ANFIS controller attains 86.5% accuracy. This disparity underscores the efficacy and potential of the novel hybrid approach, which not only adeptly handles the stochastic nature of PEMFCs but also significantly enhances accuracy in power optimization. This research not only contributes a valuable addition to the field of MPPT control but also offers a promising trajectory for the future development of efficient and reliable PEMFC systems.