Maneuver planning for autonomous vehicles, with clothoid tentacles for local trajectory planning

摘要

Autonomous navigation is nowadays a very important topic not only in the research field, but also in the industry, academic and military fields. Indeed, to accomplish this autonomy, three essential modules are needed: The perception of the environment and the vehicle localization in it, the trajectory planning and the vehicle's control. The work presented in this paper covers the trajectory planning module and can be considered as an extension of previous works presented in [1], where we developed an algorithm for local trajectory planning based on Clothoid Tentacles method. In [1], the tentacles method is used to overtake a static obstacle and turn back to the reference trajectory defined by the right lane of the road. In this paper, we aim to ameliorate the tentacles method by studying the overtaking maneuver, and generating a suitable trajectory for the lane changing maneuvers taking into account the vehicle dynamics, the road rules and some security measurements. This trajectory will serve as an intermediate reference trajectory for the vehicle in the next few seconds. The local planning algorithm is then executed with the aim of tracking the generated trajectory while avoiding possible obstacles. The maneuver planning level added to the tentacles method in this work aims to simplify the planning task and to guarantee the vehicle stability and security.