On Model Feedback Control for Robot Manipulators

Abstract

Robot manipulators are highly coupled nonlinear systems and their motions are influenced by uncertain dynamics such as coulomb friction. These nonlinearities and uncertainties disturb the performance of control systems. In this paper, a control design methodolgy is proposed for the purpose of reducing the adverse effects of parameter uncertainties and disturbances. This control structure is similar to that of classical control. Unlike classical control, this control methodology accommodates multivariate control systems with uncertain dynamics and disturbances. The control design methodology is applied to a three-degrees-of-freedom directly driven robot. Simulation and experimental results demonstrate excellent robustness.