The Trajectory Generation Method for Autonomous Vehicle Driving System

The feedforward control is a very important factor for the autonomous vehicle driving system. It is also important for this system to set the driving trajectory in the feedforward control system, which enables vehicles (e.g., a truck and bus) to follow the designated course automatically from the departure point to the destination point. This driving trajectory is the 3D geometric curve, which is generated by calculating the 3D road map data(e.g., white line marking, stop line), and includes various attributes like 3D positioning data (e.g., latitude, longitude and altitude), direction, curvature, vertical slope, cumulative distance and the position of informatory sign (e.g., stop line, cross walk). The driving trajectory with these attributes is useful for the automatic steering control system, in which the vehicle can move with safe travel speed, adequate steering angle and so on. In this paper, the authors propose the driving trajectory generation approach and demonstrate the effectiveness of it by creating the prototype software and generating each driving trajectory for experimental courses with it.