Carbon‐Aware DRL–MPC Framework for Real‐Time Control and Multiobjective Optimization of RES‐Based V2G Charging

The integration of renewable energy sources with electric vehicle (EV) charging infrastructures introduces new challenges and opportunities for modern power systems. High EV penetration leads to increased demand variability, which can compromise grid stability and elevate operational costs. Vehicle‐to‐grid (V2G) technology provides a mechanism for bidirectional energy flow, enabling EVs to not only consume electricity but also supply it back to the grid during peak periods. This paper presents a carbon‐aware real‐time control framework that integrates multiobjective optimization, deep reinforcement learning (DRL), and model predictive control (MPC) for renewable‐energy‐supported EV charging systems. The framework simultaneously minimizes operational costs, reduces carbon emissions, balances grid load, and adapts in real‐time to variations in renewable generation, EV demand, electricity prices, and carbon pricing. Extensive simulation studies demonstrate superior performance compared to baseline strategies, highlighting improvements in cost efficiency, emission reduction, and grid load management.