Enhanced Friction and Disturbance Compensation-Based Trajectory Tracking Control for Flexible Manipulator System

To deal with the problem of degradation in control performance for flexible manipulator system caused by flexible joint friction, dynamic parameter uncertainty, and external disturbance, a nonsingular fixed time sliding mode controller based on generalized disturbance estimator and adaptive radial basis function neural network is proposed in this article. First, nonlinear friction is identified by least square identification method, a generalized disturbance estimator is designed to estimate the unknown dynamic parameters and external disturbance. Nevertheless, offline identification method cannot accurately compensate the friction of flexible joints. In view of this, the friction characteristics of the flexible manipulator was analyzed, the friction identification error is approximated by the neural network. Finally, the disturbance estimation value and the friction are compensated in a feed forward manner. Compared with the traditional method that only uses offline friction compensation, this method uses radial basis function neural network to approximate friction and improves the accuracy of friction compensation. Simulation and experimental results show the advantages of proposed control method.