A Co-Simulation Framework for Assessing the Impact of Driving Styles on Energy Consumption and SOC Prediction in EVs

This study investigates the influence of different driving behaviors, namely eco-driving and aggressive driving, on energy consumption and the accuracy of the prediction of state of charge (SOC) in electric vehicles (EV). This contribution presents a co-simulation framework that integrates a physics-based vehicle model, implemented in Simulink, with a machine learning-based SOC prediction algorithm. We applied the framework to a case study assessing how different driving styles impact SOC prediction accuracy, energy efficiency, and overall energy consumption across two standard driving cycles. The findings reveal that aggressive driving results in significantly higher energy consumption, lower efficiency, and higher SOC prediction errors compared to eco-driving. These results underscore the importance of developing more robust and adaptive SOC prediction models capable of handling aggressive driving patterns.