Extreme Gradient Boosting (XGBoost) Model for Vehicle Trajectory Prediction in Connected and Autonomous Vehicle Environment

Connected and autonomous vehicles (CAVs) have the ability to receive information on their leading through multiple sensors vehicle-to-vehicle (V2V) technology then predict future behaviour thus improve roadway safety mobility. This study presents an innovative algorithm for connected determine trajectory considering surrounding vehicles. For first time, XGBoost model is developed acceleration rate that object vehicle should take based current status of both its vehicle. Next Generation Simulation (NGSIM) datasets are utilised training proposed model. The compared with Intelligent Driver Model (IDM), which a prior state-of-the-art Root Mean Square Error (RMSE) Absolute (MAE) applied evaluate two models. results show outperforms IDM in terms prediction errors. analysis feature importance reveals longitudinal position has greatest influence results.