Tracking Control of Uncertain Nonholonomic Mobile Robots: Smooth Switching Approach

Abstract

Backstepping based adaptive tracking control of non-holonomic mobile robots in the presence of both kinematic and dynamic parametric uncertainty is presented. The major challenge is the possible singularity phenomenon due to the approach of zero of the estimated input vector field entering the denominator of the control input, a common drawback of adaptive linearization-based schemes. A hybrid control approach, which switches between an adaptive and a robust control schemes, is developed for solving such a problem. It retains the advantage of an adaptive control approach to a greatest extent while avoiding the possible blowup of the torque inputs simultaneously. A case study on a specific Type (2; 0) mobile robot is provided in the final to verify the usefulness of the proposed design.