An Advanced Open-Source Model for Estimating EV State-of-Charge and Travel Time in Smart Grid Applications

Accurate estimation of electric vehicles (EVs) arrival state-of-charge (SOC) and travel time is vital for optimizing power grid operations and efficient fleet management. Existing open-source methodologies often depend on hardware and detailed data of EVs, users, routes, weather conditions, etc., limiting scalability and practical deployment. This paper presents an adaptive model that leverages real-time and publicly available datasets to dynamically predict SOC and travel time. Utilising a mean percentage error (MPE)-based learning approach, the model continuously refines its predictions by integrating historical data and real-time factors such as road conditions, traffic, temperature, and elevation changes. Performance evaluations demonstrate the model’s suitability for e-mobility stakeholders, promoting efficient resource utilisation and seamless EV fleet integration.