Supervisory traction control for a slipping UGV

Abstract

Unmanned Ground Vehicles (UGVs) face many mobility challenges as they operate in unstructured environments. In addition to insurmountable obstacles, UGVs can lose traction and slip, reducing maneuverability and potentially causing collisions with obstacles. This paper develops a dynamic model of a differentially-driven ground robot with lateral and longitudinal slip, using a ground friction model from the literature. A novel switching control system, with velocity input commands, is also presented. The controller switches from PID control to a trajectory planning mode when the robot begins to slip or when the normal command would exceed the ground friction forces. The trajectory is planned and executed based on commanded velocities. Stability and robustness of the trajectory planning controller are discussed. This control law is compared in simulation to a PID controller, using the UGV model, on different surfaces. The traction controller drives the UGV closer to the desired path derived from the reference trajectories.