Electric Vehicles Charging Sessions Classification Technique for Optimized Battery Charge Based on Machine Learning

The rapid growth in electric vehicle (EV) adoption over the last decade has increased the importance of accurately forecasting their energy consumption during charging. Lithium-ion batteries, being the primary storage units in EVs, require careful management to prevent overcharging, thereby preserving battery health and extending their lifespan. This paper introduces a Machine Learning model based on the XGBoost algorithm to forecast the duration of EV charging sessions. The model predicts the charging duration by assigning each event to its appropriate class, where each class corresponds to a specific interval of charging times. The prediction relies solely on the information available at the start of the charging event, including arrival time, initial state of charge (SOC), and calendar-related data. The model is validated using a real-world dataset comprising charging records from over 100 users, and a sensitivity analysis is conducted to evaluate the impact of different input features. Compared to benchmark models, the proposed XGBoost-based approach demonstrates superior performance, achieving a notable improvement in forecasting accuracy.