The Study On Submerged Arc Welding Inside Seam Tracking System Of Spiral Pipe

During the production of spiral, the weld seam is the typical long one and coupledwith the slight change of the helix angle, workers have to operate the welding machinefrequently, then it leads to a stronger labor intensity and the welding quality also can’t beguaranteed reliably, and even in the presence of security risks, then the automatic seamtracking system of spiral pipe is particularly important.This topic based on the tracking system of the outside weld seam, and combined withthe spiral pipe production line of Liao Yang Da Ze steel pipe plant, conducted a study onthe image processing of the inside weld seam of spiral pipe.Due to the space limitations of the welding inside the pipe, this paper uses the methodthat collects the back temperature field of the weld seam by CCD, through the temperaturefield image recognition technology to extract the position information of welding torch,then realize the automatic seam tracking program. On the hardware side, used the8mmindustrial endoscope as the CCD camera and designed a adjustable bracket to fix the CCDcamera depend on the requirement of experiment.In the software research, in order to improve the efficiency of the development andreduce the dependence on the scene, this paper improved and developed the originalsimulation player software, through using the DIB functions, it broke the mode that playvideo first and then intercept image and research algorithm by dealing with static image. Itcan read and write the image directly from the memory by the DIB functions, and thisgreatly improves the speed of image processing and basically achieves the purpose thatdealing with the image while playing, it makes the simulation play more realistic.In image processing, this paper tried to use different algorithms on the imagepreprocessing, and realized the segmentation of temperature field image by usingpiecewise linear transformation of the gray scale, it greatly facilitates the extraction of theweld seam and the edge of temperature field. Using Laplace operator as the method ofedge detection, and extracted the location information of the weld seam successfullythrough the projection method. The results show that these algorithms can identify the location information of weld seam and arc well, and the recognition time of the algorithmsis about100ms, it can meet the requirement of the seam tracking system of spiral pipe andprovides sufficient experimental data and algorithm support for the application of the entiresystem in actual production in the future.