A Practical Study of Time-Elastic-Band Planning Method for Driverless Vehicle for Auto-parking

摘要

This paper presents a practical study of Time-Elastic-Band (TEB) path planner for driverless vehicle autoparking. TEB is a dynamic optimization planner that can plan a path for a vehicle subject to vehicle's kinematic constraints. Since the basic form of the TEB algorithm cannot generate a desired trajectory for an autonomous vehicle auto-parking, we propose a two-stage TEB planning method for integrating autonomous navigation and auto-parking seamlessly. After desired auto-parking trajectory generated by TEB planner, we adopt two PID controllers (accelerator and brake) with hysteresis for vehicle velocity control. We also implement a geometric steering controller to control the steering angle of our car such that it is able to follow the generated trajectory closely. We have successfully demonstrated our approach for two driverless vehicle vertical auto-parking scenarios, namely reverse-in parking and parallel parking. Testing results are available at the following URL, https://www.youtube.com/watch?v=taEncZ-FZIA.