Optimization of EV battery charging for temperature control

Abstract

Electric vehicles (EVs) offer a sustainable mode of transportation; however, excessive battery temperature rise during charging degrades performance, limits lifespan, and affects the safe operation of power converters and EV drive systems. To address this problem, a back propagation neural network (BPNN) based temperature prediction model is integrated with four nature-inspired optimization techniques (NIOTs): whale optimization algorithm (WOA), moth flame optimization (MFO), modified particle swarm optimization (MPSO), and grey wolf optimization (GWO), to optimize multi-stage charging current profiles. Among the evaluated methods, WOA achieves the best performance, reducing charging time by approximately 10% (11400 s to 10300 s) and average temperature rise by nearly 50% (3.9 °C to 1.9 °C) compared to the conventional constant current-constant voltage (CC-CV) strategy. The optimized charging currents directly support improved DC-DC converter operation and stable EV drive performance by limiting thermal stress and current transients. Overall, accurate thermal-aware charging enhances charging efficiency, ensures safe converter operation, and contributes to reliable and long-life EV battery and drive system performance.