Research On Pipeline Straight Seam Welding And Seam Recognition Technology Of 6R Robot

With the exploitation of oil,natural gas and other resources,the demand for large-scale industrial transportation pipelines is also greatly increased.Due to their large diameter、large thickness and welds on the curved surface,these large-scale industrial transportation pipelines are difficult to weld manually and have low efficiency.Moreover,the welding quality is easily affected by the personal factors of welders.Some unqualified welding quality will lead to the scrapping of the whole pipeline,which lead to increasing the manufacturing cost.Using robot automatic welding can overcome the shortcomings of low efficiency and unstable welding quality in manual welding,and can also carry out welding operation continuously.For a single fixed workpiece,the teaching method can be used.After teaching once,the continuous welding operation can be carried out.However,in the actual manufacturing and production,the diameter,thickness and other parameters of the pipe will change.If the pipe workpiece of each size is taught once,it will be time-consuming and inefficient.Therefore,it is necessary to be able to automatically identify the weld without being affected by the size and placement position of the parts to be welded.Therefore,after studying the relevant technologies of laser vision guided weld recognition,this paper determines the monocular vision pipeline weld recognition scheme guided by line structured light,which selects the"Eye-in-hand" installation scheme of welding gun and industrial camera.Using the toolbox of MATLAB completes the calibration of internal parameters,external parameters,distortion coefficient of industrial camera and line structured light plane equation.The calibration of the welding gun tool is completed by using StaubliCalTool7 toolkit of Staubli robot,and the calibration of the hand eye relationship is completed by using OpenCV visual open source library.After calibrating of the whole weld recognition system,the solution model of three-dimensional information of weld feature points is established.In order to solve the coordinate information of weld feature points from the weld image with laser stripes collected by industrial camera,a weld image processing algorithm is developed.In order to remove the noise of the weld image,the weld image is filtered by the median filtering algorithm.The threshold segmentation of weld image is completed by using the maximum interclass variance method.According to the width of the laser stripe in the weld image,the region of interest(ROI)is extracted from the image.Based on the improved geometric center algorithm,the extraction of the center line of the single pixel skeleton of the laser stripe is realized fast and accurately.The equation of the laser stripe is fitted and calculated by the least square method.According to the characteristics of the laser stripe of the pipeline weld image breaking at the weld,establish the solution model of the weld feature points,and calculate the pixel coordinates of the weld feature points accurately.Based on the three-dimensional measurement principle of line structured light,the feature points in the weld structure are reconstructed from the pixel coordinate system to the three-dimensional coordinate system.According to the standard requirements of welding posture,the model of calculating robot welding posture is established accurately by using a series of weld feature points and multi views.Based on thickness of the pipeline to be welded and the quality of the weld,calculate the interpolation points and realize the welding path of the three typical strip transportation modes of sawtooth,triangle and crescent,which improves the quality and beauty of the weld.According to the control flow of pipeline weld recognition system guided by line structured light,using Qt and OpenCV vision library develop systems software including calibration of vision system,image processing and three-dimensional reconstruction,and interpolation point calculation and path planning of three strip transportation.The software interface is friendly.After building the experimental platform of line structured light guided weld recognition system,carry out the weld feature point recognition experiment,robot welding strip transportation experiment,welding posture verification experiment and welding experiment.The experiments show that the accuracy of the weld recognition algorithm developed in this paper is within 0.5mm,and the pose solution algorithm is accurate and feasible,and the weld quality obtained by using the three welding strip transportation methods is high and the shape is beautiful,which can meet the requirements of automatic welding of pipeline weld robot.