Application of MRAC theory for adaptive control of a constrained robot manipulator

This study presents the compliance control of a robot manipulator under a c onstrained environment. Considering the impact of frictional force allows u s to more accurately simulate the relationship between the manipulator and environmental contact forces. The controller design proposed herein is base d on the adaptive control scheme. In this design, the feed forward and feed back controllers control the position and the contact force of end-effector . Applying these controllers allows us to adapt the manipulator to the unkn own surface of the surrounding environment and to have close contact with t he curved surface. A Lyapunov function ensures the stability of the system. For an unknown contour, most controllers fail since the desired trajectory can not be obtained, and the parameters of the manipulator are unknown and may varied with the contact force and position. However, in this study, wh en performing compliant motion, the desired trajectory is generated from th e controller based on the tangential direction of the measured contact forc e. This tangential direction changes according to the operating point. We a pproach the original nonlinear system with a second order linear system at each instantaneous operating point. Correspondingly, the contour of workpie ces can be measured. Experimental results conform the feasibility of the pr oposed adaptive control scheme. Without knowledge of the contour of workpie ce and gains of the controller in advance, the adaptive controller performs well for various unknown contours. (C) 2000 Elsevier Science Ltd. All righ ts reserved.