A novel machine learning based maximum power point tracking in interleaved buck-boost converter for portable solar-powered electric vehicle charging in rural areas

Abstract

Need for off-grid electric vehicle (EV) charging solutions, using intelligent control systems, such as machine learning (ML)-based maximum power point tracking (MPPT), to harness solar energy, offers a means of optimizing efficiency even in the face of fluctuations. This innovative strategy combines clean energy, cutting-edge power electronics, and practical application, which makes it perfect for fostering innovation in areas with inadequate infrastructure. For rural areas without grid infrastructure, this paper presents a novel design and performance assessment of a portable solar-powered EV charging system. To maximize solar energy harvesting and charging efficiency, the system combines an interleaved buck-boost converter with an ML-driven MPPT algorithm. It is appropriate for small electric vehicles (EVs) like auto rickshaws because it uses a 48 V lithium iron phosphate (LiFePO₄) battery. A supervised regression model trained on real-time electrical (voltage, current, and power) and environmental (temperature, irradiance) parameters is used to implement the MPPT algorithm. The system was created using MATLAB/Simulink, and the key performance parameters were evaluated using real-time information. Analyses of the key performance metrics like charging efficiency, converter stability, and tracking accuracy show a superior energy harvesting efficiency of 97%.