In this paper a nonlinear learning control design is introduced for trajectory tracking of robot manipulators with actuator dynamics. The learning control is developed based on the Lyapunov direct method and, for repeated tasks, global stability is always guaranteed. Because the control design and the stability analysis are done directly using the nonlinear model, common assumptions and limitations used in learning control such as Lipschitz condition, acceleration measurement, and resetting of initial conditions, are not required. Furthermore, the proposed learning control is robust since exact knowledge of robot dynamics is not required except for their bounding functions. Simulation results show the effectiveness of the proposed scheme.
展开▼
