Hybrid adaptive/robust motion control of rigid-link electrically-driven robot manipulators

A learning control scheme for a class of robot manipulatorswhose endpoint is moving under geometrical constraints on a surfaceis proposed. In this scheme, the input torque command is composed oftwo different signals updated separately at every trial by different ways.One is updated by the angular velocity error vector which is projectedto the tangent plane of the constraint surface in joint space. The otheris updated by the magnitude of contact force error at the manipulatorendpoint.Not only the uniform boundedness of position and velocity trajec-tory errors but also the uniform convergence of position and velocitytrajectories to their desired ones with repeating practices are provedtheoretically. In addition, it is shown that the contact force itself convergesto the desired one in the sense of L2-norm with repeating practices.Computer simulation results by using a 3 DOF manipulator are presentedto demonstrate the effectiveness of the proposed method and to examinethe speed of convergence of force trajectories besides position and velocitytrajectories. Manuscript received April 26, 1993; revised June 10, 1994.The authors are with the Department of Mathematical Engineering andInformation Physics, Faculty of Engineering, University of Tokyo, Bunkyo-ku, Tokyo, 113 Japan.IEEE Log Number 9410118.