Research On Robotic Grinding Planning Technology Based On Point Cloud Information Of Special-Shaped Castings

In the field of aerospace manufacturing,in order to improve equipment performance and meet the needs of light weight,aerodynamics,special coatings,etc.,many parts are designed into complex shapes with various curved surfaces,but due to limited casting technology,thin-walled shaped parts are difficult to mold at one time,and a lot of grinding work on the workpiece is often required.However,the grinding and removal of the rough parts of these parts are mainly manual grinding,which has low efficiency and much harm.This topic will use the high flexibility and developability of industrial robots,and integrate the point cloud of the actual workpiece to study the grinding planning technology of shaped castings,and it is of important significance to improve the automation in the manufacturing process and the adaptability to the workpieces in the grinding process.First,based on the actual workpiece point cloud,the extraction method of the grinding area is studied.The original point cloud of the workpiece under the working environment is obtained by the laser 3D scanner,and a series of point cloud preprocessing such as environmental point cloud separation,point cloud filtering,point cloud registration,etc.are conducted to obtain the point cloud of the region of interest to be segmented,which is the input of the segmentation extraction algorithm.Aiming at the problem that the existing single point cloud segmentation method is prone to over-segmentation when dealing with poor surface quality and unevenness,a hybrid segmentation algorithm based on region growth and boundary is proposed to reduce the impact of surface abnormal points on segmentation and ensure segmentation quality.Then,base on the extracted point cloud area to be polished,a polishing planning method based directly on the point cloud data is studied,which avoids the huge calculation caused by the large amount of data in the point cloud reconstruction.Aiming at the difference between the abrasive belt grinding contact and the machine tool cutting contact,the improved residual height calculation method is adopted to change the setting of the cutting plane spacing in the planning process to improve the quality of the grinding path.In the process of path point interpolation,the interpolation method is improved based on the idea of equal curvature,so that the distribution of path points adapts to changes in curvature.The collision-free optimization of the planning results,one is to plan a collision-free path from the starting position of the tool to the starting point of grinding,and the other is to to perform collision detection on the generated path points to avoid collisions during the processing.Finally,the experimental verification of robot polishing planning technology is carried out.A grinding simulation system is built in ROS,and the planning results are imported into the simulation system to analyze the grinding motion of the robot,and then the planning results are evaluated.An experimental platform for robotic grinding and processing is built,which is to verify the proposed extraction method of the grinding area and the robotic grinding planning method,and then the robot polishes the actual workpiece surface,and analyzes the polished workpiece surface to evaluate the method described in this article.