Adaptation Dominant-Type Adaptive Learning Controller for Robot Manipulators Using a Forgetting Factor

This paper describes a new Adaptation Dominant-Type Adaptive Learning Controller (ADTALC) for trajectory tracking of robot manipulators. In this method, the torque of adaptive controller becomes much larger than the torque of learning controller when the joint angle trajectories converge to the desired ones. In this proposed controller, a forgetting factor is introduced, so that the torque of learning controller approaches zero. The stability of this proposed controller is proved by Lyapunov stability analysis, and its effectiveness is demonstrated by the computer simulation results of a 2-link robot manipulator. This proposed controller provides a novel approach for robot control system to achieve a high-speed and high-precision trajectory tracking.