Robust Tracking Controller Design With Uncertain Friction Compensation Based on a Local Modeling Approach

This paper presents a new methodology for the design of a robust controller to compensate for friction-induced dynamical characteristics inherently present in servodrives systems. A friction model is developed using a local modeling approach of the physical properties of friction along the operating range of the underlying system. Generally, developing a faithful model for physical nonlinearities is still a challenging task that is strongly related to the identification effort required by the structure of the model and the complexity of the control algorithm. The proposed model has the advantage of being simple and able to describe friction locally. The accuracy of the estimator based on the model structure can be improved by a gain-scheduled input signal obtained for different velocities and used as a precompensator of nonlinear friction. This leads to an effective linearzing strategy of the controlled system that subsequently simplifies the controller implementation stage. A stabilizing-state feedback controller is designed, assuming an inexact compensation of friction, which guarantees robustness against uncertainties arising from modeling errors and achieves high tracking performance of the overall controlled system. Experimental tests performed on a robot joint laboratory prototype demonstrate the effectiveness of the proposed friction compensation scheme to improve the performance of the overall system.