A Machine Learning Approach for EV Traffic Volume Prediction

Accelerating the adoption of Electrical Vehicles (EVs) is a key strategy to effectively transition towards decarbonized transportation systems and achieving net-zero energy goals. To this end, building a robust EV charging infrastructure plays an important role. This research is part of the broader efforts to develop CalderaCast, a web-application tool designed for EV charging stations planning. Our proposed approach involves leveraging machine learning and clustering techniques to predict the hourly traffic volumes on the nearest highways/segments of a proposed EV charging station’s location. The predicted traffic volumes will be utilized to forecast the potential station’s load behavior.