焊接缺陷磁光成像动态检测与识别

发布时间：2019-02-13 20:33

【摘要】：为了实现焊接缺陷的自动检测,研究一种交变磁场激励下焊缝表面及亚表面缺陷的磁光成像动态无损检测方法。分析了基于法拉第磁致旋光效应的焊接缺陷磁光成像机理,并结合交变磁场原理推导出励磁变化与动态磁光成像的关系。探索低碳钢板的亚表面焊缝磁光成像特征试验,验证了所提方法可用于检测焊缝亚表面的未熔合缺陷。最后对高强钢焊缝特征的动态磁光图像进行分析,采用主成分分析法和支持向量机(PCA-SVM)模式识别方法建立了焊接缺陷分类模型。试验结果表明,所提方法可以识别高强钢焊件中的焊缝特征(未熔透、裂纹、凹坑和无缺陷),缺陷分类模型的整体识别率达到92.6%,能够实现焊缝表面及亚表面缺陷的自动检测。
[Abstract]:In order to detect welding defects automatically, a dynamic nondestructive testing method for weld surface and subsurface defects under alternating magnetic field excitation was studied. The mechanism of magneto-optic imaging of welding defects based on Faraday magneto-optic effect is analyzed, and the relationship between excitation variation and dynamic magneto-optic imaging is deduced in combination with the principle of alternating magnetic field. The magneto-optic imaging test of the subsurface weld of low carbon steel plate was investigated, and it was verified that the proposed method could be used to detect the unfused defects on the subsurface of the weld. Finally, the dynamic magneto-optic image of weld characteristics of high strength steel is analyzed, and the classification model of welding defects is established by principal component analysis (PCA) and support vector machine (PCA-SVM) pattern recognition. The test results show that the proposed method can identify weld characteristics (non-penetration, cracks, pits and no defects) in high-strength steel welds, and the overall recognition rate of the defect classification model is 92.6. Automatic detection of weld surface and subsurface defects can be realized.
【作者单位】：广东工业大学机电工程学院;
【基金】：国家自然科学基金资助项目(No.51675104) 广东省科技计划资助项目(No.2016A010102015) 广州市科技计划资助项目(No.201510010089)
【分类号】：TG441.7

【相似文献】