Vibration control and angular tracking of a flexible link via neural networks

Abstract

In this paper, we employ the adaptive Neural Network (NN) control to supress the vibration a flexible robotic manipulator with a proper tip-payload. The model of the flexible manipulator system is presented with the Lumped Spring-Mass method which can enhance the accuracy in reflecting the elastic vibrations of the original system. Full-state feedback control is carried out to approximate the unmeasured variables in this paper. Uniform ultimate boundedness (UUB) of the manipulator system is achieved via the Lyapunov's direct theory. Simulations for the lumped flexible manipulator system are presented to verify the effectiveness of the proposed NN control strategies, and the control performance is compared with a PD control strategy.