Font Size: a A A

Research On Welding Seam Pouting Detection Technology Of Straight Seam Welded Pipe

Posted on:2017-03-23Degree:MasterType:Thesis
Country:ChinaCandidate:Y D ZhangFull Text:PDF
GTID:2321330536955767Subject:Control Science and Engineering
Abstract/Summary:
In this paper,we do a thorough research in the welding seam pouting detection technology of straight welded pipes.Firstly,expounds the cause of the welding seam pouting of straight welded pipes,and with the current research status at home and abroad,to determine the detection method of straight welded pipes’ defects by using laser scanning technology.Using laser scanner to get the 3d contour data of straight seam welded pipe,and then process the data,choose the method of randomized Hough transform to detect the circle to extract the characteristic value of the straight welded pipes’ defects,to detect the welding seam pouting defects of the straight seam welded pipe.In order to improve the randomized Hough transform algorithm efficiency and the accuracy,the 3d contour data should to be separated to welded pipe surface of circular arc and weld characteristics data,using the circular arc data for randomized Hough transform to extract the characteristic value of welded pipe,so propose to use k-means clustering algorithm to optimize the randomized Hough transform algorithm.Because of the characteristics of the welded pipe’s surface data,the effect of using k-means clustering on the data points directly is not very good.In order to improve the accuracy of k-means clustering algorithm,so propose to use differential curvature algorithm to optimize the k-means clustering algorithm.Use differential curvature algorithm to calculate all the curvature of the discrete points,it can be seen that the curvature of the data points on the weld and the arc data on both sides of weld are very different.Make use of this characteristic,we can use k-means clustering algorithm to divide them,and then use the randomized Hough transform to detect the circle.Through experimental analysis,the accuracy and real-time performance of improved k-means clustering algorithm and randomized Hough transform algorithm are improved obviously.Finally,based on the method of welding seam pouting detection technology of straight welded pipes in this paper,designed a welding seam pouting detection system of straight welded pipes under the environment of MATLAB/GUI.The experimental results show that the system can determine the welding seam pouting of straight seam welded pipes.
Keywords/Search Tags:Straight welded pipe, Weld pout, Randomized Hough transform, K-means clustering, Curvature
Related items