Enhancing trajectory tracking accuracy of industrial robots through temporal–spatial mapping and multi-measurement alignment

Abstract

Robotic machining and automatic offline programming has been developing rapidly over the last decade, yet the absolute accuracy of industrial robots significantly impacts the processing performance, limiting their application in high-precision manufacturing fields. For dynamic non-contact continuous robotic machining tasks, such as laser cutting, precise trajectory tracking performance is especially critical. In this paper, a parallel tracking error compensation framework is proposed to improve the tracking performance of industrial robots, based on temporal–spatial mapping and multi-measurement alignment (TSM-MMA). Major tracking errors originate from nonlinear motor control lag and non-geometric motion transmission errors. The proposed method incorporates both servomotor encoder feedback and laser tracker measurements, enabling parallel distinction and compensation of these errors across trials. Typical linear and circular trajectories are analyzed using TSM to normalize multi-sensor data. Gaussian process regression (GPR) is employed in the MMA process to model the regularity of repetitive measurements, facilitating targeted error compensation. Physical experiments are conducted with an EFORT ER14-1400 robot and a Leica AT960 laser tracker to validate the effectiveness of the proposed framework.