基于反卷积神经网络的超声导波层合板脱粘缺陷超分辨成像

doi:10.16180/j.cnki.issn1007-7820.2023.08.002

摘要/Abstract

摘要：

针对传统超声导波成像检测方法难以精确表征结构损伤细节特征的问题,文中提出了基于深度学习的反卷积神经网络模型,对层合板中亚波长脱粘缺陷的超分辨成像问题进行研究,以获得损伤的细节特征。通过有限元仿真与全聚焦成像算法获取初始成像结果,再使用数据增强方法扩充数据库,最后对标注好的12 550张损伤图像进行训练和测试。研究结果表明,与原始全聚焦成像算法相比,反卷积神经网络模型下损伤的成像位置准确度提高了5%,成像精度高于91%,定位误差低于1.8 mm,说明文中所提方法能够明显提高网络成像结果分辨率并较好地显现亚波长损伤的细节特征。上述结果表明文中所提方法具有较高的检测效率,且无需人工经验,在工程实践中具有良好的应用价值。

关键词: 层合板, 损伤检测, 有限元, 全聚焦成像, 亚波长, 脱粘缺陷, 反卷积神经网络, 超分辨率成像

Abstract:

Traditional ultrasonic guided wave imaging detection methods are difficult to accurately characterize structural damage details. In order to obtain detailed features of the damage, the deconvolutional neural network model via deep learning is proposed to investigate the super-resolution imaging problem of subwavelength debonding defects in laminate plates. Initial imaging results are obtained by finite element simulation with the total focusing method. The labeled 12 550 damage images are trained and tested based on extended database expanded by data enhancement method. The results show that compared with the original full-focus imaging algorithm, the deconvolution neural network model improves the accuracy of damage location by 5%, the imaging accuracy is higher than 91%, and the positioning error is lower than 1.8 mm, indicating that the proposed method can significantly improve the resolution of network imaging results and better display the details of subwavelength damage. The above results show that the proposed method has high detection efficiency and does not require manual experience, and has good application value in engineering practice.

Key words: laminate, damage detection, finite element, total focusing method, subwave length, debonding defects, deconvolutional neural network, super resolution

中图分类号:

TP183

岳圣尧,许伯强,徐桂东,徐晨光,张赛. 基于反卷积神经网络的超声导波层合板脱粘缺陷超分辨成像[J]. 电子科技, 2023, 36(8): 7-13.

YUE Shengyao,XU Baiqiang,XU Guidong,XU Chenguang,ZHANG Sai. Super-Resolution Imaging of Laminate Debonding Defects via Deconvolutional Neural Network and Ultrasound Guided Waves[J]. Electronic Science and Technology, 2023, 36(8): 7-13.

图/表 10

图1

表1

图2

图3

表2

图4

图5

表3

图6

图7

参考文献 25

[1]	Thamrin R, Zaidir Z, Desharma S. Debonding failure analysis of reinforced concrete beams strengthened with CFRP plates[J]. Polymers, 2021, 13(16):1-20. doi: 10.3390/polym13010001
[2]	王梓尧, 严刚. 基于Lamb波的复合材料结构分层损伤扩展监测研究[J]. 复合材料科学与工程, 2020(7):20-26.
	Wang Ziyao, Yan Gang. Monitoring of delamination growth for composite structure by using lamb waves[J]. Composite Materials and Engineering, 2020(7): 20-26.
[3]	李明轩, 王小民, 安志武. 粘接界面特性的超声检测与评价[J]. 应用声学, 2013(3):190-198.
	Li Mingxuan, Wang Xiaomin, An Zhiwu. Ultrasonic testing and evaluation of bonding interface properties[J]. Applied Acoustics, 2013(3):190-198.
[4]	包乌云. 微波无损检测消声瓦贴质量中若干问题的研究[D]. 上海: 华东师范大学, 2003:5-13.
	Bao Wuyun. Microwave nondestructive testing of anechoic tile quality in a number of issues[D]. Shanghai: East China Normal University, 2003:5-13.
[5]	骆英, 李峰忠, 徐晨光. Lamb波频域逆散射全聚焦成像方法研究[J]. 实验力学, 2021, 36(2):195-204.
	Luo Ying, Li Fengzhong, Xu Chenguang. Research on frequency-domain inverse scattering total focus method based on Lamb wave[J]. Journal of Experimental Mechanics, 2021, 36(2):195-204.
[6]	杜晗, 许桢英. 超声导波焊缝检测中阵列式传感器的信号融合方式研究[J]. 电子科技, 2020, 33(7):32-36.
	Du Han, Xu Zhenying. The research of ultrasonic wave signal fusion on array sensors in weld detection[J]. Electronic Science and Technology, 2020, 33(7):32-36.
[7]	沈成业, 洪朝, 黄海军, 等. 焊缝缺陷的全聚焦相控阵成像检测[J]. 无损检测, 2020, 42(9):45-49.
	Shen Chengye, Hong Chao, Huang Haijun, et al. Weld defects detection using TFM phased array imaging technique[J]. Nondestructive Testing, 2020, 42(9):45-49.
[8]	Zhang H, Liu Y, Fan G, et al. Sparse-TFM imaging of Lamb waves for the near-distance defects in plate-like structures[J]. Metals, 2019, 9(5):1-13. doi: 10.3390/met9010001
[9]	He J, Yuan F. A quantitative damage imaging technique based on enhanced CCRTM for composite plates using 2D scan[J]. Smart Materials and Structures, 2016, 25(10): 1-11.
[10]	Zhou K, Zheng Y, Zhang J, et al. A reconstruction-based mode separation method of Lamb wave for damage detection in plate structures[J]. Smart Materials and Structures, 2019, 28(3):1-10.
[11]	Born M, Wolf E. Principles of optics-electromagnetic theory of propagation, interference and diffraction of light[M]. Cambridge: Cambridge University Press, 1999: 461-465.
[12]	Small A, Ilev I, Chernomordik V, et al. Enhancing diffraction-limited images using properties of the point spread function[J]. Optics Express, 2006, 14(8):3193-3203. pmid: 19516461
[13]	Gruber F K, Marengo E A, Devaney A J. Time-reversal imaging with multiple signal classification considering multiple scattering, between the targets[J]. Journal of the Acoustical Society of America, 2004, 115(6):3042-3047. doi: 10.1121/1.1738451
[14]	Labyed Y, Huang L J. Ultrasound time-reversal music imaging with diffraction and attenuation compensation[J]. IEEE Transactions on Ultrasonics Ferroelectrics and Frequency Control, 2012, 59(10): 2186-2200.
[15]	Singh R, Wu W, Wang G, et al. Artificial intelligence in image reconstruction: The change is here[J]. Physica Medica-European Journal of Medical Physics, 2020, 79(2):113-125.
[16]	Shende P, Pawar M, Kakde S. A brief review on: MRI images reconstruction using GAN[C]. Chennai: Proceedings of the International Conference on Communication and Signal Processing, 2019:139-142.
[17]	Ben Yedder H, Cardoen B, Hamarneh G. Deep learning for biomedical image reconstruction: A survey[J]. Artificial Intelligence Review, 2021, 54(1):215-251. doi: 10.1007/s10462-020-09861-2
[18]	Shen Q, Yue S, Lu W, et al. Ultrasonic guided wave damage detection method for stiffened plates based on deep learning[C]. Xi'an: Proceedings of the International Conference on Intelligent Control, Measurement and Signal Processing and Intelligent Oil Field, 2021:1-6.
[19]	梁波, 卢军, 曹阳. 基于改进U-Net卷积神经网络的钢轨表面损伤检测方法[J]. 激光与光电子学进展, 2021, 58(2):334-340.
	Liang Bo, Lu Jun, Cao Yang. Rail surface damage detection method based on improved U-Net convolutional neural network[J]. Laser & Optoelectronics Progress, 2021, 58(2):334-340.
[20]	Umehara K, Ota J, Ishida T. Application of super-resolution convolutional neural network for enhancing image resolution in chest CT[J]. Journal of Digital Imaging, 2018, 31(4):441-450. doi: 10.1007/s10278-017-0033-z pmid: 29047035
[21]	Ioffe S, Szegedy C. Batch normalization: Accelerating deep network training by reducing internal covariate shift[C]. Lille: Proceedings of the International Conference on Machine Learning, 2015:1-11.
[22]	Goodfellow I, Bengio Y, Courville A. Deep learning[M]. Cambridge: MIT Press, 2016:326-366.
[23]	Zeiler M D, Taylor G W, Fergus R, et al. Adaptive deconvolutional networks for mid and high level feature learning[C]. New York: Proceedings of the IEEE International Conference on Computer Vision, 2011:2018-2025.
[24]	Wang Z, Chen J, Hoi SCH. Deep learning for image super-resolution: A survey[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(10): 3365-3387. doi: 10.1109/TPAMI.2020.2982166
[25]	周雷, 徐桂东. 基于超声逆散射模型的损伤定量成像研究[J]. 电子测量技术, 2019, 42(19):1-5.
	Zhou Lei, Xu Guidong. Research on quantitative imaging of damage based on ultrasonic inverse scattering model[J]. Electronic Measurement Technology, 2019, 42(19):1-5.

	钢板	硬橡胶板
杨氏模量/GPa	200.00	0.14
密度/ kg·m^-3	7 850.00	1 200.00
泊松比	0.30	0.49
自由四面体网格大小/ mm	0.8
时间步长/μs	0.1
检测时间/μs	150.0

数据集	损失值	精度值
训练集	0.08 734	0.931
验证集	0.05 126	0.957