Dynamic Path Following Control of a Ground Ackerman Steering Robot to Avoid a Collision

A novel method is proposed to dynamically control the path following of a ground Ackerman steering robot to avoid a collision. The method consists of collision prediction module, collision avoidance module and global path following module. The elliptic repulsive potential field method (ER-PFM) and the enhanced vector polar histogram method (VPH+) based on the Ackerman steering model are proposed to predict the collision in a dynamic environment. The collision avoidance is realized by the proposed cost function and speed control law. The global path following process is achieved by pure pursuit. Experiments show that the robot can fulfill the dynamic path following task safely and efficiently using the proposed method.