Design methodology for fully dynamic-controlled polynomial profiles and reduced tracking error in CNC machines

Abstract

High speed and high precision on computer numerically controlled (CNC) machines are major issues for most industrial automation processes. Among the most important factors for precise motion control are the control law, the controller implementation, and the profile reference. The effects of the control law and controller implementation on the tracking error in CNC machines have been widely studied, while the issue of profile reference has not received the required attention. To improve the dynamics provided by the conventional profile reference, several techniques have been proposed, focusing on peak jerk limitation. The novelty of this work is to develop a generalized methodology designed for producing dynamically constrained higher degree piecewise polynomial profiles, without limits in maximum polynomial degrees. They lead to define in an arbitrary way the dynamics shape, while the peak dynamics values constrained are maintained in order to improve the tracking error on CNC systems and giving the possibility to design fully dynamic-controlled trajectories. For experimentation purposes, a two-axis proportional–integral–derivative controller and profile reference generator are implemented in a low-cost field programmable gate array. The experimentation demonstrates the efficiency of the proposed methodology, showing the peak jerk and tracking error reduction, compared against recent reports from literature.