光学 精密工程, 2020, 28 (5): 1046, 网络出版: 2020-11-06   

交变/旋转磁场下焊接缺陷磁光成像检测与分类

Detection and classification of welding defects by magneto-optical imaging under alternating/rotating magnetic field
作者单位
广东工业大学 广东省焊接工程技术研究中心, 广东 广州 510006
摘要
针对任意角度焊接缺陷难以检测的问题, 研究在不同磁场激励下焊接缺陷磁光成像无损检测系统。重点介绍了由U形磁轭产生的交变磁场和平面交叉磁轭产生的旋转磁场激励焊件的机理, 比较了交变/旋转磁场激励下不同焊接缺陷的磁光成像效果。基于法拉第旋转效应分析磁光成像特性与磁场强度之间的关系, 磁光图像的灰度值可以匹配相应的漏磁场强度。采用主成分分析法提取融合图像列像素灰度特征和通过灰度共生矩阵提取磁光图像纹理特征, 建立BP神经网络模型和支持向量机模型识别这些缺陷特征。试验结果表明, 在旋转磁场激励下, BP神经网络模型和支持向量机模型的分类精度分别为94.1%和98.6%, 相比交变磁场, 分类精度分别提高了10.7%和8.5%。旋转磁场激励下的磁光成像克服了定向检测的局限性, 能够实现对任意角度焊接缺陷的检测及分类。
Abstract
Aiming at the difficult to detect arbitrary-angle weld defects, a Magneto-Optical (MO) imaging Non-Destructive Testing (NDT) system for weld defects excited by different magnetic fields was studied. The mechanism of the alternating magnetic field generated by the U-shaped yoke and the rotating magnetic field produced by the plane cross yoke was introduced. The MO imaging effects of different weld defects excited by alternating/rotating magnetic field were compared. The relationship between imaging characteristics of MO images and magnetic field strength was analyzed based on the Faraday rotation effect. The gray value of MO image can match the corresponding leakage magnetic field strength. The principal component analysis method was used to extract the grayscale features of the fused image column pixels and the texture features of the MO image were extracted by the gray-level co-occurrence matrix. A BP neural network model and a support vector machine model were established to identify these defect features. Experimental results show that the classification accuracy of the BP neural network model and the support vector machine model can reach 94.1% and 98.6% respectively under the excitation of rotating magnetic field. Compared with the alternating magnetic field, the classification accuracy is improved by 10.7% and 8.5%, respectively. MO imaging under rotating magnetic field excitation overcomes the limitation of directional detection of MO imaging under traditional magnetic field excitation, and realizes the detection and classification of arbitrary-angle weld defects.

李彦峰, 高向东, 季玉坤, 王春草. 交变/旋转磁场下焊接缺陷磁光成像检测与分类[J]. 光学 精密工程, 2020, 28(5): 1046. LI Yan-feng, GAO Xiang-dong, JI Yu-kun, WANG Chun-cao. Detection and classification of welding defects by magneto-optical imaging under alternating/rotating magnetic field[J]. Optics and Precision Engineering, 2020, 28(5): 1046.

本文已被 4 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!