An Inverse Dynamics-Based Discrete-Time Sliding Mode Controller for Robot Manipulators

Abstract

In the past years an extensive literature has been devoted to the subject of motion control of rigid robot manipulators. Many approaches have been proposed, such as feedback linearization [1], model predictive control [2], as well as sliding mode or adaptive control [3], [4], [5]. The basic idea of feedback linearization, known in the robotic context as inverse dynamics control [6], [7], is to exactly compensate all the coupling nonlinearities in the dynamical model of the manipulator in a first stage so that a second stage compensator may be designed based on a linear and decoupled plant. Although global feedback linearization is possible in theory, in practice it is difficult to achieve, mainly because the coordinate transformation is a function of the system parameters and, hence, sensitive to uncertainties which arise from joint and link flexibility, frictions, sensor noise, and unknown loads. This is the reason why the inverse dynamics approach is often coupled with robust control methodologies [1].