Modified Data-Driven Disturbance Observer With Its Application in Quadrant Glitch Compensation of Ball Screw Feed-Drive System

A ball screw feed-drive system (BSFDS) is a high-precision tracking and positioning mechanism with extensive application, and its performance is strongly affected by nonlinear friction. Moreover, its tracking accuracy is severely limited by quadrant glitch, which is caused by highly nonlinear behaviors with hysteresis, and are difficult to model completely. Therefore, in this paper, a data-driven three-degree-of-freedom (3-DOF) control strategy that integrates a modified disturbance observer (DOB) to compensate for quadrant glitch is proposed. First, a data-driven control method is developed with the advantage of avoiding modeling difficulties. Specifically, several structural improvements are made in the DOB to improve the system stability and quadrant glitch rejection ability. Moreover, the stability and performance of the modified DOB are analyzed. The theoretical analysis reveals that the filter in the modified DOB is less restrictive and highly tunable than that in the traditional DOB. The control strategy is experimentally verified by comparison with the feedback controller, 2-DOF controller, and traditional DOB. The results show the better quadrant glitch compensation effectiveness and adaptation of the proposed approach, improving the tracking accuracy in the BSFDS.