Development of Trajectory Prediction Methods for Oncoming Vehicles and Path Planning Techniques for Narrow Road Passing in Autonomous Driving

Abstract

We are developing autonomous driving technology for highways, general roads, and residential roads. On residential roads, cooperative actions with other vehicles, such as merging and passing, are necessary. This study proposes a learning-based method for predicting the trajectory of oncoming vehicles and a path planning method that includes reverse evasion. We report the results of testing this method in a road environment modeling residential streets.