Machine-Tool Tracking Error Reduction in Complex Trajectories Through Anticipatory ILC

High precision motion control is a major issue in machine tools. Part producing with high accuracy in reduced time requires more and more challenging capabilities of these machines. 

However at some stage the velocity and acceleration requirements get in conflict with the dynamic capabilities of the machines and the requirements have to be limited to maintain the tracking precision. The classical feedback control jointly with common feed-forward loops provides correct tracking of the commanded trajectories up to certain velocities/accelerations. Precisely, velocity and acceleration feed-forward loops eliminate the tracking error at constant speed and at constant acceleration respectively. The feedback loop reduces the deviations due to the effect of disturbances. However, depending on the machine dynamics trajectory commands need to be smoothed by lowering the jerk in order to limit/control the tracking error so as the desired accuracy is achieved, sacrificing overall trajectory time in the process. This paper proposes combining these techniques with anticipatory Iterative Learning Control (ILC) techniques to overcome that limitation when performing repetitive trajectories/parts. Moreover it can also be used to enhance the response to the not considered non-linearities and to periodic disturbances affecting the system (such as cutting forces).