Research On Online Measurement Technology Of Robot Assembly And Welding Process Based On Point Cloud Data

In the process of aircraft fuselage manufacturing,in order to improve production efficiency and reduce costs,robot laser welding technology and automatic assembly technology are widely used to assemble large thin-walled components.As a typical large thin-walled component,titanium alloy ribbed wall plate has the problems of difficult precise positioning of the component,easy deformation during welding,resulting in large assembly errors,and difficulty to meet the requirements of weld forming quality.The online realtime measurement of the stiffened plate positioning parameters before welding and the weld forming size after welding is an important means to ensure welding quality and improve automation level.The traditional two-dimensional vision measurement method is difficult to meet the requirements of robot automatic assembly and welding.The research on online real-time measurement technology of the stiffened plate assembly and welding process parameters by using three-dimensional vision is of importance.To obtain the parameters of the assembly and welding process to realize automatic robot assembly and welding,in this paper,point cloud data of the ribbed wall plate are collected by the surface structured light three-dimensional scanner,and studies the relevant algorithm of point cloud processing,to solve the problem of automatic measurement of the stiffened plate positioning parameters before welding and the weld forming size after welding.The measurement results are used to correct the assembly error and evaluate the weld forming quality,to improve the assembly accuracy and welding quality of the ribbed wall plate.The main research contents are as follows:(1)The online measurement system and experimental platform for the assembly and welding process are built by using the scanner and robot,and the scanner is calibrated to obtain the point cloud data of a single view of the workpiece.Point cloud data of complete workpiece surface cannot be collected by a single scan,a multi-viewpoint cloud stitching method based on circular marker points is adopted.Extract and match the marker points,and calculate the transformation matrix.(2)The obtained workpiece point cloud data is large and jumbled,including outliers,scene point cloud,and redundant point cloud.In this paper,the point cloud is preprocessed by statistical filtering,Euclidean clustering extraction,pass-through filtering,and the voxel grid method.A point cloud simplification algorithm based on feature extraction is proposed,which can not only meet the uniform sampling of flat areas but also preserve the weld and boundary features of the workpiece.In order to eliminate the influence of the coordinate system on the measurement of the stiffened plate positioning parameters,an alignment method of workpiece point cloud and CAD digital-analog coordinate system based on local features is proposed.The secondary development of UG to extract the local feature patch of the CAD model.and sample the patch to obtain the CAD model point cloud.(3)Before welding,the stiffened plate positioning parameters need to be measured.Firstly,the region growing algorithm is used to segment the point cloud of the stiffened plate and the wall plate.The random sample consensus(RANSAC)algorithm is used to fit the plane point cloud of the stiffened plate and the wall plate,and the plane normal vector is obtained.Then,the edge points of the stiffened plate point cloud are obtained by using the boundary extraction algorithm based on normal,and the left and right edge points are fitted into straight lines by the RANSAC algorithm.Finally,the intersection of the two straight lines and the wall plate and the average distance from the lower edge point to the wall plate are calculated.(4)After welding,the weld forming size needs to be measured.Firstly,the fast point feature histogram(FPFH)algorithm is used to extract the weld point cloud.Then the weld point cloud is sliced,and search out the two endpoints of the weld section contour point.Finally,the distance from the two endpoints to the intersection line of the stiffened plate and the wall plate and the included angle between the normal vector of the stiffened plate and the wall plate is calculated.(5)According to the requirements,the online measurement application software for the assembly and welding process is developed,and the software port is designed so that the software can exchange data with the scanner and robot.The online measurement system of the assembly and welding process is used to measure the workpiece in the workshop environment.The results show that the measurement accuracy can meet industrial requirements,can replace manual measurement,and has the advantages of non-contact,high efficiency,and so on.The stiffened plate positioning parameters before welding and the weld forming size after welding can be measured online and in real-time.