Enhancing electric vehicle battery performance through Grey Wolf Optimization and Deep Reinforcement Learning integration

This paper introduces a novel approach for optimizing the charging-discharging schedules of electric vehicle batteries by integrating Grey Wolf Optimization (GWO) with Deep Reinforcement Learning (DRL). This hybrid method leverages GWO's global search capability alongside DRL's ability to learn optimal policies on the fly, leading to enhanced battery management. Based on a dynamic battery model with standard State of Charge (SOC) representation and constraints, the optimization process prioritizes maximizing charge duration while maintaining vigilant energy usage. The results demonstrate that the proposed GWO-DRL framework achieves a 25% improvement in charging efficiency and reduces discharge energy loss by up to 30% compared to traditional methods. These results suggest the potential for significant increases in battery lifespan and dri ving range in electric vehicles while also increasing the efficiency of its operations.