Multi-sensor Data-driven Collaborative Scheduling Optimization for Electric Vehicles

In this study, we developed a collaborative scheduling model for electric vehicles (EVs) in smart microgrids by integrating charging pile operational monitoring data, photovoltaic (PV) output sensor data, and grid state sensing information.To address the operational optimization challenges of microgrids with high distributed energy penetration, we established an EV scheduling model.The system dynamics were characterized using a Markov decision process framework, and the proximal policy optimization algorithm was applied to optimize strategies that target maximum PV utilization.Research findings indicate that the coordinated charging of a growing EV fleet, guided by price response patterns, serves to absorb excess solar power and enhance PV utilization rates.These results validate the pivotal role of multimodal sensor data fusion in enabling the flexible regulation of power systems, while providing methodological support for optimizing sensor network deployment in the Internet of Energy Things.