Trajectory predictions of lane changing vehicles using SVM

Abstract

In this paper, we present Support Vector Machine (SVM) based prediction of the trajectory of a lane changing vehicle. SVM is used for both short–range and long–range trajectory prediction of lane changing vehicles. A lane change manoeuvre is potentially dangerous and erroneous estimation may cause a collision. The necessity of forewarning drivers of the feasibility of a safe lane change requires forecast of vehicle trajectories. For this, trajectories of vehicles involved in the lane change are modelled as discrete time series using actual field data. Results indicate that SVM is able to forecast the lane change trajectory with sufficient accuracy.