Composite adaptive control of flexible joint robots

Abstract

A composite adaptive controller is developed for flexible joint robots with inertia parameter uncertainty. It does not require any restrictions on the joint stiffness, since a full-order dynamic model is used in the design and stability analysis. Previously published results based on the fourth-order model require at least joint jerk feedback and derivatives of the manipulator regressor (up to the second-order). In contrast, the proposed adaptive controller requires at most joint acceleration feedback. Its adaptive law is of the same complexity as the well-known Slotine and Li's algorithm for rigid-body robots. Asymptotic stability of the adaptive controller is established in the Lyapunov sense. Simulation results are included to demonstrate the tracking performance.