Adaptive control of robotic servo system with friction compensation

Abstract

Adaptive control design is studied for a class of robotic turntable servo systems with friction dynamics. A newly developed continuously differentiable friction model is utilized, where the explicit model parameters are incorporated into the neural network weights adaptation. To further enhance the adaptation convergence and to reduce the tracking error, a velocity predictor is proposed to develop a composite adaptation law combining the tracking error and the predictor error. Stability analysis and bounds for both errors are guaranteed, and extensive experimental investigations are performed on a two-axis robotic turntable to assess the advantages and the achievable control performance of the proposed schemes.