Command-based iterative learning control with TCP function for five-axis contouring

Abstract

This paper presents a novel approach for precision machining in five-axis machine tools through the use of five-axis command-based iterative learning control (FCILC). A command-based iterative learning control algorithm with tool center point (TCP) control and interpolation function is developed to improve the tracking and contouring performance of the TCP for the repetitive contoured trajectories. The TCP control function is utilized to generate smooth velocity profiles and interpolation commands where the jerk of each axis is limited by the given constraints. The contour error estimation method for five axes is developed to efficiently compute the contour errors. Applying the FCILC technique to the original control architecture enables reducing tracking error caused by servo lag and friction, and alleviating noise effect during repetitive learning tasks. Finally, contouring experiments are conducted on a five-axis engraving machine to demonstrate that the tracking and contour errors can be enhanced significantly using the proposed FCILC method.