Deep Reinforcement Learning-Based Energy Management for Renewable-Powered EV Charging Stations

The increasing adoption of electric vehicles (EVs) and renewable energy sources has introduced significant challenges in the real-time management of EV charging stations. This paper proposes an intelligent energy management system based on the Deep Deterministic Policy Gradient (DDPG) algorithm for a photovoltaic-powered EV charging station equipped with a battery energy storage system. To improve decision-making under dynamic operating conditions, an Adaptive Multi-Objective Reward Function (AMRF) is developed to simultaneously minimize electricity costs, reduce grid dependency, maximize renewable energy utilization, and satisfy EV charging demands. The proposed framework is evaluated under various charging scenarios and compared with a conventional rule-based strategy. The results demonstrate improved operational efficiency and enhanced renewable energy utilization, highlighting the effectiveness of the proposed approach.

کلیدواژه ها:

Electric Vehicles ، Deep Reinforcement Learning ، DDPG ، Energy Management System ، Renewable Energy ، Battery Energy Storage System ، Smart Charging ، Smart Grid