激光超声表面缺陷检测的过程中，缺陷的定量表征通常依赖于操作者的判断，易受到人为因素干扰，致使检测结果不稳定。针对这一问题，提出一种基于图像识别的二维卷积神经网络（2D-CNN）的缺陷自动分类检测方法。利用有限元方法模拟激光超声检测过程，并采集超声信号数据用于训练分类模型；使用连续小变换（CWT）处理超声信号得到小波时频图，以小波时频图作为输入训练卷积神经网络（CNN）分类模型，实现对表面缺陷深度的自动分类。验证结果表明：提出的检测方法能够对不同深度的缺陷准确分类，测试的平均准确率达到97.3%；构建的CNN分类模型能够自主学习输入图像的缺陷特征并完成分类，提高了检测结果稳定性，为激光超声缺陷检测的自动化分析处理提供了新的思路。
Abstract
In the process of the laser ultrasonic surface defects detection, the quantitative characterization of the defects mainly depends on the operator's judgment, and it is easily interfered by the human factors, which leads to the unstable detection results. To solve this problem, an defects automatic classification detection method based on the two-dimensional convolutional neural network (2D-CNN) for image recognition was proposed. The finite element method was used to simulate the laser ultrasonic detection process, and the ultrasonic signal data was collected for training the classification model; the continuous wavelet transformation (CWT) was used to process the ultrasonic signal to obtain the wavelet time-frequency images, and the images were used as inputs to train the convolutional neural network (CNN) classification model to realize the automatic classification of the surface defects depth. The verification results show that the proposed detection method can accurately classify the defects of different depths, and the average accuracy rate of the test reaches 97.3%; the constructed CNN classification model can independently learn the defects features of the input images and complete the classification, which improves the stability of the test results, and provides a new idea for the automatic analysis and processing of laser ultrasonic defects detection.
关键词
缺陷分类 /
连续小波变换 /
卷积神经网络 /
激光超声检测
{{custom_keyword}} /
Key words
laser ultrasonic detection /
convolutional neural network /
defects classification /
continuous wavelet transformation
{{custom_keyword}} /
基金
大连理工大学中央高校基本科研业务费专项项目（DUT15ZD110）
{{custom_fund}}
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
周正干, 孙广开, 李洋. 先进无损检测技术在复合材料缺陷检测中的应用[J]. 航空制造技术, 2016(4): 30-35.
黄燕杰, 尚建华, 任立红, 等. 用于铝板缺陷无损检测的激光超声有限元模拟研究[J]. 应用光学,2019,40(1):150-156.
陶程, 殷安民, 王煜帆, 等. 激光激发表面波测量表面缺陷深度的数值研究[J]. 激光与红外,2019,49(1):42-50.
施成龙, 师芳芳, 张碧星. 利用深度神经网络和小波包变换进行缺陷类型分析[J]. 声学学报,2016,41(4):499-506.
郑远攀, 李广阳, 李晔. 深度学习在图像识别中的应用研究综述[J]. 计算机工程与应用,2019,55(12):20-36.
李萍, 宋波, 毛捷, 等. 深度学习在超声检测缺陷识别中的应用与发展[J]. 应用声学,2019,38(3):458-464.
袁建虎, 韩涛, 唐建, 等. 基于小波时频图和CNN的滚动轴承智能故障诊断方法[J]. 机械设计与研究,2017,33(2):93-97.
董晗, 孔超, 师芳芳, 等. 超声相控阵
SUN G K, ZHOU Z G, LI G H, et al. Development of an optical fiber-guided robotic laser ultrasonic system for aeronautical composite structure testing[J]. Optik,2016,127(12):5135-5140.
ZHOU Zhenggan, SUN Guangkai, LI Yang. Application of advanced nondestructive testing technology in composite material defect detection[J]. 2016(4): 30-35.
EVERTON S, DICKENS P, TUCK C, et al. Evaluation of laser ultrasonic testing for inspection of metal additive manufacturing[J].Proceeding of SPIE, 2015, 9353: 935316.
YANG L, UME I C. Inspection of notch depths in thin structures using transmission coefficients of laser-generated Lamb waves[J]. Ultrasonics,2015,63:168-173.
HERNANDEZ-VALLE F, DUTTON B, EDWARDS R S. Laser ultrasonic characterisation of branched surface-breaking defects[J]. NDT & E International,2014,68:113-119.
HUANG Yanjie, SHANG Jianhua, REN Lihong, et al. Finite element simulation in laser ultrasound for non-destructive testing of aluminum defect materials[J]. Journal of Applied Optics,
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
