Abstract
The authors address the problem of controlling rigid robot manipulators subject to bounded torque input disturbances and to joint velocity measurements bounded noise. The advantages of the control scheme presented are mainly that global convergence is guaranteed without prior knowledge of the bounds on the plant parameters and disturbances. If these bounds are known a priori, they can be used to improve the control scheme performance. Disturbances on the plant input as well as on the plant output are considered. The cases discussed are for when the disturbances are bounded and the upperbounds may be known or unknown. and for when the disturbances are linear with respect to some unknown parameters.<>