Abstract
The adaptive motion control of robotic manipulators is investigated using a deterministic design approach, which requires only a knowledge of bounds on the system uncertainty. Given a nominal model of the manipulator dynamics, the feedback control scheme consists of an approximate feedback linearizing control followed by a robust controller incorporating an on-line adaptive algorithm. Robust trajectory tracking is shown under certain norm inequalities. The theory is illustrated by simulating the first three joints of PUMA 560 robot arm.