Ultrasonic detection method for weld defects of longitudinal submerged arc welded pipe based on multi-scale U-Net

Abstract

The invention relates to a multi-scale U-Net-based ultrasonic detection method for weld defects of a longitudinal submerged arc welded pipe. The ultrasonic detection method overcomes the problem of low accuracy of weld defect detection of the longitudinal submerged arc welded pipe under a complex background in the prior art. The method provided by the invention can effectively improve the accuracy of detecting the weld pipe defect of the longitudinal submerged arc welded pipe under the complex background. The method comprises the following steps: step 1, meshing a longitudinal submerged arc welded pipe weld ultrasonic image by adopting a spectral clustering algorithm and distributing labels to obtain a super-pixel image; step 2, calculating a gray average value of all pixel points in each superpixel, and re-assigning the gray average value to each pixel point in the superpixels to obtain a visual summary graph; step 3, constructing an improved multi-scale U-Net model; 4, the obtained visual summary graph serves as a training set to train an improved multi-scale U-Net model, and the trained multi-scale U-Net model is used for detecting the weld defect of the longitudinal submerged arc welded pipe.

Description

Technical field: [0001] The invention belongs to the technical field of welding seam defect detection of welded pipes, and relates to an ultrasonic inspection method for welding seam defects of straight seam submerged arc welded pipes based on spectral clustering and multi-scale U-Net fusion. Background technique: [0002] There may be various defects in the weld of straight seam submerged arc welded pipe. Welding defects not only reduce the cross-sectional area of ​​the pipeline, but also tend to cause stress concentration, and may even induce brittle fracture. In particular, there is a notch effect at the tip, which is prone to three-way stress state, resulting in crack instability and expansion, resulting in structural fracture, which may cause serious accidents and major economic losses, endangering personal safety. How to detect the weld defects of LSAW pipes early and accurately has always been an important research direction. There are many detection methods, but the...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to provide a multi-scale U-Net-based ultrasonic detection method for weld defects of longitudinal submerged arc welded pipes, which overcomes the limitations of the detection of weld defects of longitudinal submerged arc welded pipes under the complex background existing in the prior art The problem of low accuracy

Images