Torque Feedforward Control of the Parallel Spindle Head Feed Axes

The hybrid machine tool based on parallel spindle heads can efficiently process complex parts, but the load of the feed axes of the spindle head has the characteristics of nonlinear and coupling, which brings difficulties to the control. In this paper, the contour error of the tool center point (TCP) is greatly reduced by the method of dynamic modeling and feedforward. Specifically, the kinematic modeling is performed using the vector loop method, the dynamic modeling is performed using the virtual work principle, and the frictional force modeling is performed using the Stribeck model to obtain the dynamic model of the spindle head. The servo system of the feed axis is modeled and then the torque feedforward module is designed. Finally, in MATLAB/Simulink simulation, torque feedforward effectively reduces the contour error of TCP. Based on the analysis of the simulation results, the load torque of the parallel spindle head has the characteristics of nonlinearity and coupling. The evaluation index LCE (load caused error) is proposed, and the load torque mainly causes errors during acceleration, deceleration and commutation.