A Multi-agent Autonomous Formation Motion Planning Method for Car-Like Robots Based on the Leader-Follower Model

With the continuous application and development of unmanned technology in the military field, unmanned combat systems have gradually become an important force in modern warfare and even future warfare. The purpose of this paper is to study the coordinated formation movement planning of multiple unmanned wheeled combat vehicles under the background of focusing on the development of the future war situation. The typical formation framework of the leader-follower structure is used to combine quadratic programming and polynomial planning methods to realize the trajectory planning of formation members, and the decoupling planning is carried out based on the Frenet coordinate system to reduce the dimension and simplify the problem. The strategy of virtual vehicles guiding followers is proposed to ensure the stable driving of the formation. Finally, experiments are carried out on the Gazebo simulation platform and the real environment, which proves the effectiveness of the method.