Fuel cells are emerging as a clean and efficient power source for electric vehicles (EV). But, due to their nonlinear characteristic nature, they are failed to maintain maximum efficiency. Standard MPPT algorithms often converge slowly and fluctuate around the maximum power point, which lowers overall energy use. To address these problems, this study proposes a modified long short-term memory (MLSTM) based controller for maximum power point tracking. This method uses predictive learning to estimate the maximum power point accurately and create adaptive control signals for the DC–DC converter. This ensures stable fuel cell operation and effective power delivery to the traction system. Simulation results demonstrate that the proposed approach achieved a tracking time of 0.57 s, THD of 9.66%, and efficiency of 99.25%, outperforming the conventional P&O method. The proposed controller not only improves the overall energy efficiency of the fuel cell system but also extends the driving range of the EV with renewable energy sources in hybrid vehicle architectures. This topology provides a promising solution to fuel cell based EV in terms of enhanced sustainability and efficiency.
Preethi et al. (2025) studied this question.