Evaluation of three model reference adaptive control algorithms for robotic manipulators

Abstract

This paper evaluates three model reference adaptive control (MRAC) schemes for completing a telerobotic task. The three proposed schemes are evaluated for accurate trajectory control of a general three-degree-of-freedom robotic manipulator, in the presence of large payload variations, and modeling inaccuracies. The three MRAC schemes evaluated are: (1) Computed-torque method, which uses the nonlinear dynamic model of the robot in the control formulation and Popov's hyperstability criteria, (2) independent joint control method, which uses decoupled linear dynamic equations in the control formulation and based on Popov's hyperstability criteria, and (3) independent joint control method based on sensitivity analysis. Computer simulations of a three-degree-of-freedom manipulator, with a large payload and fast maneuver are used to analyze the performance of the three schemes. The investigation shows the robustness of the computed-torque method when compared to the independent joint schemes.