关键词: 叶片打磨, 平面分割, 叶片点云分割, 轨迹规划

Abstract: Aiming at the problem of poor point cloud segmentation effect in the grinding process of aviation engine blades and the large amount of calculation caused by surface fitting in trajectory planning, a point cloud based method for aircraft engine blade segmentation and grinding trajectory planning was proposed, aiming to improve point cloud segmentation effect and reduce the amount of calculation. The point cloud of the blade to be polished was collected by the structured light camera, pre-processed such as filtering was carried out, and the Random Sample Consensus (RANSAC) plane segmentation algorithm was improved to improve the efficiency of the algorithm. A region growing segmentation algorithm combined with point cloud color information was designed to achieve stable segmentation of the blade area to be polished. A triple section method directly based on point cloud was proposed, combined with B-spline curve fitting algorithm to generate pre-grinding trajectory, and based on equal residual height algorithm to optimize line spacing, complete grinding trajectory planning. It has been verified by experiments that the efficiency of the plane segmentation algorithm is 40.25 % higher than that of the traditional method, and the data processing efficiency is greatly improved. In addition, through the improved region growing segmentation algorithm, the efficiency of blade grinding trajectory planning is significantly improved, and its realization effect is significantly enhanced.

Key words: blade grinding, plane segmentation, blade point cloud segmentation, trajectory planning