A Hybrid Model Reference Adaptive Control/Computed Torque Control Scheme for Robotic Manipulators

Abstract

A direct model reference adaptive controller scheme for a robotic manipulator based on the Lyapunov stability criterion is developed. The algorithm presented is based on the formulation of a plant state model that includes all the non-linear characteristics of the robot manipulator. This adaptive controller takes advantage of the structure and any known dynamics of the system in order to increase the speed of adaptation and relax the conditions required for convergence. It uses a modified Lyapunov function to derive the adaptation laws that are not dependent on the usual assumption that the time-varying plant is close to the desired time-invariant plant. The Lyapunov design guarantees asymptotic stability.