Feedforward combined multi-axis cross-coupling contour control compensation strategy of optical mirror processing robot

Abstract

During the movement of an optical mirror processing robot (OMPR), the movement error of each branch chain leads to contour errors of the grinding tool, which reduce the accuracy of the optical mirror surface. To improve the processing accuracy of an OMPR, it is necessary to study the control and compensation strategy of its contour error. In this study, first, a kinematics analysis of an OMPR is conducted, and the trajectory of the end execution point in the world coordinate system is transformed into the fixed coordinate system of the robot. Combined with the common trajectory of optical mirror processing, based on the Frenet coordinate system, contour error models of the OMPR in straight line, arc, and spiral trajectories are established. Subsequently, the contour error, feedforward channel gain, and compensation channel gain models of the parallel module are established in the task space, and concurrently, the control variables and stability of the system are analyzed. Finally, the established feedforward combined multi-axis cross-coupling contour control compensation strategy is analyzed experimentally to verify its accuracy and effectiveness. It provides a theoretical basis for a robot to directly face the precision processing object using the control and compensation strategy in a future research study to improve the molding accuracy of a surface and optimize the processing technology of a large-scale optical mirror.