Reinforcement Learning for Intelligent Vehicle-to-Grid Integration: Balancing Clean Energy Utilization and Battery Degradation Preservation

The transition toward sustainable mobility requires an intelligent integration of Electric Vehicles (EVs) into the power grid. This paper proposes a Reinforcement Learning (RL) framework using Proximal Policy Optimization (PPO) to manage bidirectional Vehicle-to-Grid (V2G) power flow. Utilizing the Indian Grid Master dataset, the system optimizes for economic arbitrage and carbon reduction while strictly adhering to a 90% State of Charge (SoC) mobility requirement. A core innovation of this work is an asymmetric reward function that applies a 35x penalty to battery discharge relative to charging, ensuring hardware longevity. Results across various 10-hour shift profiles demonstrate the agent's ability to achieve mobility targets while maximizing grid stability.