Optimizing Electric Vehicle Charging Considering Driver Satisfaction Through ML Algorithm

Abstract

The rapid adoption of electric vehicles (EVs) has increased the demand for efficient and intelligent charging infrastructure. However, long waiting times, station congestion, and varying electricity prices often reduce driver satisfaction during charging sessions. To address these challenges, this study proposes a machine learning–based optimization system for electric vehicle charging that considers driver satisfaction as a primary factor. The system is implemented using a Django-based web framework integrated with advanced machine learning models to analyze charging conditions and predict driver satisfaction levels. The proposed system collects multiple parameters related to EV charging, including state of charge before charging, queue length, waiting time, electricity price, distance to charging station, station rating, charger power, historical satisfaction, price fairness, and charging speed stability, along with contextual factors such as weather conditions, crowd level, preferred charging time, charging station type, and trip urgency. These features are processed using data preprocessing techniques such as standardization, categorical encoding, and feature alignment before training machine learning models. Multiple algorithms, including Random Forest, Gradient Boosting, and XGBoost, are trained and evaluated to determine the most effective model for predicting driver satisfaction levels. The system compares models using performance metrics such as accuracy, precision, recall, and F1-score, and selects the bestperforming model for deployment. Once trained, the system predicts the expected satisfaction level of a driver during a charging session and provides intelligent recommendations to improve the charging experience, such as selecting less crowded stations, choosing off-peak charging times, or using faster chargers. By integrating machine learning prediction with real-time user interaction through a web interface, the proposed system helps drivers make informed charging decisions and assists service providers in optimizing charging station operations. Overall, the proposed approach improves charging efficiency, reduces waiting time, and enhances driver satisfaction, contributing to a smarter and more sustainable EV charging ecosystem.