Approximate Filippov-based Switching for Behavior-based Trajectory Tracking
Ryan Jacobson, Greg N. Droge · 2021
This paper presents a switching strategy based upon first-order Filippov switching conditions for safely following a trajectory that passes through obstacles. The Filippov switching conditions motivate a sliding mode behavior to bridge the gap between the two control modalities. Thus, a novel wall-following controller is developed that keeps pace with the trajectory while the trajectory temporarily passes through obstacles. The products of this work are a three-mode automaton and vector field behaviors that can be used to simultaneously track the desired trajectory while avoiding obstacles. This paper's findings were used to successfully control a TurtleBot3 Burger robot in a Gazebo simulation.