Page 75 - 无损检测2024年第十期
P. 75
李长侑,等:
基于深度学习和时间反演的复合材料内扩展性损伤重建
探讨了使用U-Net直接反演方法在尽量少的天线数 4366-4376.
量并且天线采用直线型排列的情况下的损伤重建效 [9] FINK M.Time reversal of ultrasonic fields.I.basic
果。最后,基于这两种方法,提出了TR-Unet方法并 principles[J]. IEEE Transactions on Ultrasonics,
进行了应用效果验证,结果表明,该方法不仅可以满 Ferroelectrics,and Frequency Control,1992,39(5):
555-566.
足实际工程需求,还在形状相似度和轮廓细节的准
[10] CARMINATI R,PIERRAT R,DE ROSNY J,et al.
确性上取得了显著提升。
Theory of the time reversal cavity for electromagnetic
参考文献: fields[J]. Optics Letters,2007,32(21):3107-3109.
[11] CHEN Y M,WANG B Z.Polycentric spatial focus
[1] CASTELLANO A,FRADDOSIO A,PICCIONI M D.
of time-reversal electromagnetic field in rectangular
Quantitative analysis of QSI and LVI damage in GFRP conductor cavity[J]. Optics Express,2013,21(22):
unidirectional composite laminates by a new ultrasonic
26657-26662.
approach[J]. Composites Part B:Engineering,2018,151:
[12] MUKHERJEE S,TAMBURRINO A,HAQ M,et al.
106-117.
Far field microwave NDE of composite structures using
[2] 雷毅. 无损检测技术问答[M]. 北京:中国石化出版社,
time reversal mirror[J]. NDT & E International,2018,
2013.
93:7-17.
[3] WANG Y M,CHEW W C.An iterative solution of
[13] MUKHERJEE S,MAYS R O,TRINGE J W.A
the two-dimensional electromagnetic inverse scattering
microwave time reversal algorithm for imaging extended
problem[J]. International Journal of Imaging Systems and
defects in dielectric composites[J]. IEEE Transactions on
Technology,1989,1(1):100-108.
Computational Imaging,2021,7:1215-1227.
[4] VAN DEN BERG P M,KLEINMAN R E.A contrast
[14] 刘彻,杨恺乔,鲍江涵,等. 智能电磁计算的若干进
source inversion method[J]. Inverse Problems,1997,
展[J]. 雷达学报,2023,12(4):657-683.
13(6):1607-1620.
[15] WEI Z,CHEN X D.Deep-learning schemes for full-
[5] CHEN X D.Subspace-based optimization method
wave nonlinear inverse scattering problems[J]. IEEE
for solving inverse-scattering problems[J]. IEEE
Transactions on Geoscience and Remote Sensing,2019,
Transactions on Geoscience and Remote Sensing,2010,
57(4):1849-1860.
48(1):42-49.
[16] ZHANG H H,YAO H M,JIANG L J,et al.
[6] JIN K H,MCCANN M T,FROUSTEY E,et al.
Solving electromagnetic inverse scattering problems in
Deep convolutional neural network for inverse problems
in imaging[J]. IEEE Transactions on Image Processing: inhomogeneous media by deep convolutional encoder‒
a Publication of the IEEE Signal Processing Society, decoder structure[J]. IEEE Transactions on Antennas
2017,26(9):4509-4522. and Propagation,2023,71(3):2867-2872.
[7] YAO H M,JIANG L J,NG M.Implementing the fast [17] SONG R C,LI M L,XU K W,et al.Electromagnetic
full-wave electromagnetic forward solver using the deep inverse scattering with an untrained SOM-net[J]. IEEE
Transactions on Microwave Theory and Techniques,
convolutional encoder-decoder architecture[J]. IEEE
Transactions on Antennas and Propagation,2023,71(1): 2022,70(11):4980-4990.
1152-1157. [18] SALADI P,KALEPU Y.Electromagnetic inverse
[8] LIN Z C,GUO R,LI M K,et al.Low-frequency data scattering problem solved by DConvNet and adapted
prediction with iterative learning for highly nonlinear attention U-net[C]//2023 IEEE 12th International
inverse scattering problems[J]. IEEE Transactions on Conference on Communication Systems and Network
Microwave Theory and Techniques,2021,69(10): Technologies(CSNT).Bhopal,India: IEEE,2023.
41
2024 年 第 46 卷 第 10 期
无损检测

