Non-Linear Adaptive Control of Robots Including Motor Dynamics

Abstract

This paper addresses the theoretical aspects of non-linear adaptive control of robots when motor dynamics is included. It is shown that the resulting third-order dynamic models can be transformed into an attractive form, under moderate assumptions, to which the most recent controller design methods are applicable. The third-order model retains important properties (properties 1, 2 and 3 in the paper) which are present in the second-order rigid-body models of robots. Control methods are proposed to tackle two main issues: parameter adaptation for unknown dynamic parameters and robustness in the sense of modelling errors. It is shown that the control methods for rigid-body robots can easily be extended for higher order systems that possess properties 1, 2 and 3.