Energy Consumption of Electric Buses Under Operational Data of Transit System

Paper in proceeding, 2024

Urban adoption of electric buses prompts a need for optimized energy management. This study addresses the over-looked influence of operational data on battery electric buses (BEBs) energy consumption. It investigates which operational parameters significantly impact BEBs' energy usage and determines the most predictive data features. Employing Random Forest (RF) and eXtreme Gradient Boosting (XGBoost), the study develops regression models evaluated by RMSE, MSE, MAE, and R-squared metrics. Results indicate robust predictive performance with RF and XGBoost achieving RMSEs of 1.88 and 1.75, respectively. Euclidean distance between trip origins and destinations is identified as the key factor influencing energy consumption. The research enhances route planning and operational efficiency for BEBs, offering significant implications for reducing energy costs and supporting data-driven decisions in transit management.