Sustainable Electric Vehicle Charging Station Energy Management using Model Predictive Control and Ensemble Learning

The rapid increase in electric vehicles (EVs) requires smarter infrastructure for efficient and sustainable charging. Managing diverse energy sources at EV charging stations calls for a sophisticated energy management system (EMS) capable of balancing supply and demand while minimizing operational costs. This paper presents an EMS that employs ensemble learning for precise prediction of photovoltaic power generation and EV charging demand. By utilizing a two-level hierarchical Model Predictive Control (MPC) approach, the system optimizes power allocation, ensuring all connected EVs are sufficiently charged without overburdening the infrastructure. This integrated approach effectively addresses uncertainties in renewable energy production and fluctuating EV demand, promoting efficient and reliable energy management for EV charging stations.