A MATHEMATIC APPROACH TO ELECTRIC VEHICLE TRAVEL RANGE PREDICTION USING IOT AND MACHINE LEARNING WITH ROAD CONDITIONS AND CROWDSOURCED CHARGING DATA

Abstract

Electric vehicle (EV) is increasingly promoted for their environmental and economic benefits but still challenges of range prediction and reliable access to charging infrastructure in long trip journey is necessary. Estimating driving range is complex due to its dependent on multiple factors such as traffic condition, internal load, charging patterns, road conditions, driving behavior and environmental conditions. In this paper proposes a hardware and machine learning collaborative system to predict the EV travel range prediction by incorporating road conditions data. In the proposed system various onboard sensors collect information about road conditions like flat, uphill, downhill with battery parameters which are then processed using ML driven analysis to estimate the travel distance based on the available battery charge and road conditions. The system also allows other users to analyze the road condition for their upcoming trip based on historical available data and the system recommends a travel range. Additionally, the system integrates with IoT enabled charging stations to provide real time updates on station availability along with crowdsourcing mechanism verified by user whether the charging station is public or private. By combining predictive analytics IoT connectivity and user driven data sharing this integrated approach improves trip planning reduces the risk of vehicle stranding and increases charging accessibility for EV users.