基于深度学习的焊缝PAUT数据智能化分析方法

doi:10.13700/j.bh.1001-5965.2020.0578

北京航空航天大学学报 ›› 2022, Vol. 48 ›› Issue (3): 504-513.doi: 10.13700/j.bh.1001-5965.2020.0578

基于深度学习的焊缝PAUT数据智能化分析方法

朱甜甜^1,2, 宋波¹, 毛捷^1,2, 廉国选¹

1. 中国科学院声学研究所声场声信息国家重点实验室, 北京 100190;
2. 中国科学院大学电子电气与通信工程学院, 北京 100049

收稿日期:2020-10-12 发布日期:2022-03-29
通讯作者: 宋波 E-mail:songbo@mail.ioa.ac.cn
基金资助:
船舶建造焊缝质量数字化检测技术研究项目

PAUT data intelligent analysis method of welding seams based on deep learning

ZHU Tiantian^1,2, SONG Bo¹, MAO Jie^1,2, LIAN Guoxuan¹

1. State Key Laboratory of Acoustics, Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, China;
2. School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

Received:2020-10-12 Published:2022-03-29
Supported by:
Research on Digital Inspection Technology for Ship Construction Weld Quality

摘要/Abstract

摘要： 超声相控阵检测技术(PAUT)凭借其突出的技术优势被广泛应用在船舶、铁路、石油石化和航空航天等诸多领域。在焊缝超声相控阵检测(PAUT)中，对检测数据缺陷的识别定位目前多采用传统的人工判读方式，判读效率较低，对检测人员的判读经验有较高要求，难以满足自动化超声检测的要求。基于深度学习中的目标检测和跟踪算法构建智能识别模型，通过对焊缝超声相控阵检测的S、B扫图特征进行融合，并结合焊缝的三维结构信息，识别并定位出缺陷在焊缝中的三维空间位置。实验结果显示:缺陷框的平均三维IOU(预测三维缺陷框和实际三维缺陷框的平均交并比)达到0.644 9，较为接近缺陷的真实空间位置，可以实现焊缝超声相控阵检测成像结果智能识别和定位。

关键词: 超声相控阵检测(PAUT), 焊缝检测, 深度学习, 目标检测, 跟踪算法, 缺陷识别, 三维定位

Abstract: In the welding seam phased array ultrasonic testing (PAUT), the traditional manual judgment method is used to identify and locate the defects in the inspection data. However, this method has lower interpretation efficiency and higher requirements for the experience of the inspectors, and it has difficulty for meeting the requirements of automated ultrasound inspection. In this paper, combined with the features of S and B scan images of welding seam PAUT and 3D structure of weld, an intelligent recognition model based on target detection and tracking algorithm in deep learning is proposed to identify and locate the weld defect automatically. The experimental results show that the average value of 3D IOU of the defects (the average intersection ratio of the predicted and the actual 3D defect frame) reaches 0.644 9, which is close to the real defects’ location. This method can realize the intelligent recognition and positioning from PAUT imaging data in welding seam.

Key words: phased array ultrasonic testing (PAUT), weld seam testing, deep learning, target detection, tracking algorithm, defect recognition, three-dimensional positioning

中图分类号:

TP399

朱甜甜, 宋波, 毛捷, 廉国选. 基于深度学习的焊缝PAUT数据智能化分析方法[J]. 北京航空航天大学学报, 2022, 48(3): 504-513.

ZHU Tiantian, SONG Bo, MAO Jie, LIAN Guoxuan. PAUT data intelligent analysis method of welding seams based on deep learning[J]. Journal of Beijing University of Aeronautics and Astronautics, 2022, 48(3): 504-513.

参考文献

[1] 李京安.超声相控阵扇形扫描检测成像技术研究[D].哈尔滨:哈尔滨工业大学,2009.LI J A.Research on phased array ultrasonic sector-scan inspection imaging technology[D].Harbin:Harbin Institute of Technology,2009(in Chinese).
[2] 陈彦华,李明轩.人工神经网络在超声无损检测中的应用[J].应用声学,1996,15(3):40-44.CHEN Y H,LI M X.Application of artificial neural network to ultrasonic nondestructive testing[J].Applied Acoustics,1996,15(3):40-44(in Chinese).
[3] 刘镇清.人工神经网络BP算法的改进及其在无损检测中的应用[J].测控技术,2001,20(3):56-58.LIU Z Q.Improvement of BP training algorithm for artificial neural network and its application in NDT[J].Measurement & Control Technology,2001,20(3):56-58(in Chinese).
[4] 陈国华,张新梅,谢常欢,等.超声检测中裂纹型缺陷深度的智能识别[J].华南理工大学学报(自然科学版),2005,33(8):1-5.CHEN G H,ZHANG X M,XIE C H,et al.Intelligent recognition of crack depth in ultrasonic testing[J].Journal of South China University of Technology (Natural Science),2005,33(8):1-5(in Chinese).
[5] 黄民,李功.焊缝超声无损检测中的缺陷智能识别方法[J].北京信息科技大学学报,2009,24(2):33-36.HUANG M,LI G.Intelligent defect recognition methods in the ultrasonic non-destructive test of welds[J].Journal of Beijing Information Science and Technology University,2009,24(2):33-36(in Chinese).
[6] 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.SHI C L,SHI F F,ZHANG B X.Analysis on defect classification by deep neural networks and wavelet packet transform[J].Acta Acustica,2016,41(4):499-506(in Chinese).
[7] MENG M,CHUA Y J,WOUTERSON E,et al.Ultrasonic signal classification and imaging system for composite materials via deep convolutional neural networks[J].Neurocomputing,2017,257:128-135.
[8] 黄焕东,胡利晨,李斌彬,等.基于区域的快速卷积神经网络的焊缝TOFD检测缺陷识别[J].无损检测,2019,41(7):12-18.HUANG H D,HU L C,LI B B,et al.Recognition of defect in TOFD image based on faster region convolutional neural networks[J].Nondestructive Testing Technologying,2019,41(7):12-18(in Chinese).
[9] 刘志浩.对接焊缝的超声相控阵检测及三维成像分析[D].南昌:南昌航空大学,2016.LIU Z H.Butt welding inspection and defects three-dimensional imaging using ultrasonic phased array technology[D].Nanchang:Nanchang Hangkong University,2016(in Chinese).
[10] REN S Q,HE K M,GIRSHICK R,et al.Faster R-CNN:Towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,39(6):1137-1149.
[11] GIRSHICK R.Fast R-CNN[C]//Proceedings of the IEEE International Conference on Computer Vision (ICCV).Piscataway:IEEE Press,2015:1440-1448.
[12] BEWLEY A,GE Z Y,OTT L,et al.Simple online and realtime tracking[C]//2016 IEEE International Conference on Image Processing (ICIP).Piscataway:IEEE Press,2016:3464-3468.
[13] 包俊,董亚超,刘宏哲.基于Deep-Sort的目标跟踪算法综述[C]//中国计算机用户协会网络应用分会2019年第二十三届网络新技术与应用年会,2019:13-17.BAO J,DONG Y C,LIU H Z.Survey of object tracking algorithms based on Deep-Sort[C]//China Computer Users Association,Proceedings of the 23rd Annual Conference on New Network Technologies and Applications,2019:13-17(in Chinese).
[14] 闫庆森,李临生,徐晓峰,等.视频跟踪算法研究综述[J].计算机科学,2013,40(S1):204-209.YAN Q S,LI L S,XU X F,et al.Survey of visual tracking algorithm[J].Computer Science,2013,40(S1):204-209(in Chinese).
[15] 尹宏鹏,陈波,柴毅,等.基于视觉的目标检测与跟踪综述[J].自动化学报,2016,42(10):1466-1489.YIN H P,CHEN B,CHAI Y,et al.Vision-based object detection and tracking:A review[J].Acta Automatica Sinica,2016,42(10):1466-1489(in Chinese).
[16] 彭丁聪.卡尔曼滤波的基本原理及应用[J].软件导刊,2009,8(11):32-34.PENG D C.Basic principle and application of Kalman filter[J].Software Guide,2009,8(11):32-34(in Chinese).
[17] 王学斌,徐建宏,张章.卡尔曼滤波器参数分析与应用方法研究[J].计算机应用与软件,2012,29(6):212-215.WANG X B,XU J H,ZHANG Z.On analysis and application approach for Kalman filter parameters[J].Computer Applications and Software,2012,29(6):212-215(in Chinese).
[18] 叶斌,徐毓.强跟踪滤波器与卡尔曼滤波器对目标跟踪的比较[J].空军雷达学院学报,2002,16(2):17-19.YE B,XU Y.Comparison between a strong tracking filter and Kalman filter for target tracking[J].Journal of Air Force Radar Academy,2002,16(2):17-19(in Chinese).
[19] 周鑫,何晓新,郑昌文.基于图像深度学习的无线电信号识别[J].通信学报,2019,40(7):114-125.ZHOU X,HE X X,ZHENG C W.Radio signal recognition based on image deep learning[J].Journal on Communications,2019,40(7):114-125(in Chinese).

基于深度学习的焊缝PAUT数据智能化分析方法

PAUT data intelligent analysis method of welding seams based on deep learning

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	黄晨威, 程景春, 潘雄, 宋凝芳, 刘冰. 基于深度神经网络的像素级别可见光图像配准[J]. 北京航空航天大学学报, 2022, 48(3): 522-532.
[2]	刘春辉, 王美琳, 董赞亮, 王沛. 基于调制卷积神经网络的空地数据链信道估计[J]. 北京航空航天大学学报, 2022, 48(3): 533-543.
[3]	沈元, 严寒冰, 夏春和, 韩志辉. 一种基于深度学习的恶意代码克隆检测技术[J]. 北京航空航天大学学报, 2022, 48(2): 282-290.
[4]	兰凌强, 刘淇缘, 卢树华. 基于注意力机制与特征相关性的人脸表情识别[J]. 北京航空航天大学学报, 2022, 48(1): 147-155.
[5]	包壮壮, 赵学军. 基于EfficientDet的无预训练SAR图像船舶检测器[J]. 北京航空航天大学学报, 2021, 47(8): 1664-1672.
[6]	韩永赛, 马时平, 何林远, 李承昊, 朱明明, 许悦雷. 改进的深度神经网络下遥感机场区域目标检测[J]. 北京航空航天大学学报, 2021, 47(7): 1470-1480.
[7]	陆婷婷, 张尧, 阎岩, 杨利民, 杨卫东. 一种基于自动特征学习的陨石坑区域检测算法[J]. 北京航空航天大学学报, 2021, 47(5): 939-952.
[8]	陈科山, 郝宇, 何泓波, 李坤龙. 基于R-D SSD模型航空发动机安装工位检测算法[J]. 北京航空航天大学学报, 2021, 47(4): 682-689.
[9]	陈智能, 史存存, 李轩涯, 贾彩燕, 黄磊. 一种时空特征聚合的水下珊瑚礁鱼检测方法[J]. 北京航空航天大学学报, 2021, 47(3): 509-519.
[10]	徐晓华, 钱平, 王一达, 周昕悦, 徐汉麟, 徐李冰. 面向电力系统的多粒度隐患检测方法[J]. 北京航空航天大学学报, 2021, 47(3): 520-530.
[11]	李昊, 管荑, 王杉, 石玮, 刘子鑫, 刘晓川. 电力系统厂站接线图拓扑关系检测技术[J]. 北京航空航天大学学报, 2021, 47(3): 531-538.
[12]	李昊, 王杉, 耿玉杰, 王黎, 孙文昌, 苗纯源. 基于深度学习和图匹配的接线图检测与校核[J]. 北京航空航天大学学报, 2021, 47(3): 539-548.
[13]	谢彭宇, 徐新. 基于多尺度联合学习的行人重识别[J]. 北京航空航天大学学报, 2021, 47(3): 613-622.
[14]	马腾宇, 李孜, 刘日升, 樊鑫, 罗钟铉. 基于无监督学习的多模态可变形配准[J]. 北京航空航天大学学报, 2021, 47(3): 658-664.
[15]	王新, 李喆, 张宏立. 一种迭代聚合的高分辨率网络Anchor-free目标检测方法[J]. 北京航空航天大学学报, 2021, 47(12): 2533-2541.