Tracking Controllers for Robot Manipulators: A High Gain Perspective

In this paper we propose a controller for the path tracking problem of robotic manipulators. The proposed controller structure is very similar to that of the "computed-torque" scheme widely known in the literature. However, unlike the "computed-torque" scheme this controller is robust with respect to parameter uncertainty and disturbances. The robustness is provided by a high gain mechanism that works in conjunction with a nonlinear precompensator in the control loop. The fundamental action of the high gain is to minimize the effective uncertainty that remains, subsequent to any cancellation by the precompensator. The synthesis of this controller is quite easily formulated and its performance is dictated primarily by a single high-gain parameter, thus making it particularly attractive for on line tuning. Some guidelines are provided for the off-line selection of this high-gain parameter.