Low Velocity Friction Compensation and Feedforward Solution Based on Repetitive Control

Abstract

One of the most significant sources of tracking error for an X-Y table is static friction, a nonlinear disturbance at low velocity. In traversing a circular profile, an X-Y table encounters zero velocity crossings at ninety degree intervals around the circle, leaving relatively large tracking errors referred to as “quadrant glitches.” To learn the control input which eliminates errors caused by stiction, repetitive control, a subclass of learning control, is employed. Experiments, conducted on the X-Y bed of a CNC milling machine, demonstrate that near-perfect tracking can be achieved in twelve cycles or less. Of particular interest is the velocity command input generated by repetitive control. The input contains feedforward information for developing perfect trajectories at low velocities.