Dynamic carbon footprint electric vehicle charging network intelligent scheduling system based on reinforcement learning and blockchain collaborative optimization

The large-scale popularisation of electric vehicles (EVs) is confronted by two major challenges. Firstly, there is the need to reduce the carbon emissions of the transport sector. Secondly, there are fluctuations in grid load. This study proposes an experimental simulation of public data in the California region of the United States. The simulation is under the dual action of a reinforcement learning algorithm that optimises charging scheduling strategies in real time and a framework transaction mechanism that combines blockchain technology to build a transparent and trusted dynamic smart charging network. The experimental results demonstrate the efficacy of the framework in balancing the load on the power grid, enhancing the efficiency of distributed charging, and reducing costs and increasing efficiency (reducing carbon emissions by approximately 5%–15% in comparison with conventional scheduling) while ensuring data ownership. This study provides a robust theoretical foundation for the utilisation of reinforcement learning algorithms and blockchain technology to advance the transformation of intelligent transportation and ESG goals.