As urban infrastructure becomes denser, subway operations face numerous risks, including adjacent construction and structural damage deterioration. Integrated monitoring is crucial to ensure safety. This study presents a multitask vibration transformer (M-TVT) model, which utilizes an Expert-Gate network to manage task-related information sharing. The model achieves integrated identification for the type and spatial distribution of internal and external tunnel abnormal vibration responses, along with assessing structural damage degree. Meanwhile, a tunnel-train-soil coupling experiment is conducted to validate the efficacy of the M-TVT model. The results demonstrate that the M-TVT model with the Expert-Gate network maintains high accuracy and stability, efficiently shares parameter information among tasks and achieves improvement in accuracy compared to single-task models.
PengFLQiaoYKYangC.Building a knowledge graph for operational hazard management of utility tunnels. Expert Syst Appl2023; 223: 119901.
2.
YuXYChenZLiuF, et al. How urban metro networks grow: from a complex network perspective. Tunnelling Underground Space Technol2023; 131: 104841.
3.
MooneyMAGormanPBGonzalezJN.Vibration-based health monitoring of earth structures. Struct Health Monit2005; 4: 137–152.
4.
FanWQiaoPZ.Vibration-based damage identification methods: a review and comparative study. Struct Health Monit2011; 10: 83–111.
5.
HoshyarANSamaliBLiyanapathiranaR, et al. Structural damage detection and localization using a hybrid method and artificial intelligence techniques. Struct Health Monit2020; 19: 1507–1523.
6.
BiaoZXiongyaoXXiaojianW.The tunnel structural mode frequency characteristics identification and analysis based on a modified stochastic subspace identification method. Shock Vib2018; 2018: 6595841.
7.
BiaoZFengshouZXiongyaoX.Vibration characteristics of underground structure and surrounding soil underneath high speed railway based on field vibration tests. Shock Vib2018; 2018: 3526952.
8.
WangZYiTHYangDH, et al. Early warning method of structural damage using localized frequency cointegration under changing environments. J Struct Eng2023; 149: 04022230.
9.
LiHXieXZhangY, et al. Theoretical, numerical, and experimental study on the identification of subway tunnel structural damage based on the moving train dynamic response. Sensors2021; 21: 7197.
10.
FlahMNunezIBen ChaabeneW, et al. Machine learning algorithms in civil structural health monitoring: a systematic review. Arch Comput Method Eng2021; 28: 2621–2643.
11.
AlazzawiOWangDS.A novel structural damage identification method based on the acceleration responses under ambient vibration and an optimized deep residual algorithm. Struct Health Monit2022; 21: 2587–2617.
12.
AbbasNUmarTSalihR, et al. Structural health monitoring of underground metro tunnel by identifying damage using ANN deep learning auto-encoder. Appl Sci Basel2023; 13: 1332.
13.
AlazzawiOWangDS.Health monitoring of an assembly-type structure using raw time-domain EMI signals and a novel hybrid deep learning algorithm. Smart Mater Struct2024; 33: 025038.
14.
ChaiFZhouBXieX, et al. Incremental learning model for dynamical identification and classification of abnormal vibration in operational underground facilities. Tunnelling Underground Space Technol2023; 140: 105251.
15.
ZhangYXieXLiH, et al. An unsupervised tunnel damage identification method based on convolutional variational auto-encoder and wavelet packet analysis. Sensors2022; 22: 2412.
16.
ZhangYCYiTHLinSB, et al. Automatic corrosive environment detection of RC bridge decks from ground-penetrating radar data based on deep learning. J Perform Constr Facil2022; 36: 04022011.
17.
LiYTBaoTFHuangXJ, et al. An integrated underwater structural multi-defects automatic identification and quantification framework for hydraulic tunnel via machine vision and deep learning. Struct Health Monit2023; 22: 2360–2383.
18.
VaswaniAShazeerNParmarN, et al. Attention is all you need. arXiv:1706.03762.
19.
CardenEPFanningP.Vibration based condition monitoring: a review. Struct Health Monit2004; 3: 355–377.
20.
KoabazMHajzargarbashiTKunduT, et al. Locating the acoustic source in an anisotropic plate. Struct Health Monit2012; 11: 315–323.
21.
EntezamiAShariatmadarHMarianiS.Fast unsupervised learning methods for structural health monitoring with large vibration data from dense sensor networks. Struct Health Monit2020; 19: 1685–1710.
22.
ZhouBGuiYWangX, et al. Three-level distributed real-time monitoring of construction near underground infrastructure using a combined intelligent method. Sensors (Basel)2022; 22: 3260.
23.
ZhangYLXieXYLiHQ, et al. Tunnel damage identification method based on relative entropy of wavelet packet energy: an experimental verification. In: IOP Conference Series: Earth and Environmental Science, Beijing, China, 23–26 October 2020, vol. 570, p. 032034.
24.
ChenBHuTYHuangZS, et al. A spatio-temporal clustering and diagnosis method for concrete arch dams using deformation monitoring data. Struct Health Monit2019; 18: 1355–1371.
25.
MarinielloGPastoreTMennaC, et al. Structural damage detection and localization using decision tree ensemble and vibration data. Comput Aided Civ Inf2021; 36: 1129–1149.
26.
KordSTaghikhanyTAkbariM.A novel spatiotemporal 3D CNN framework with multi-task learning for efficient structural damage detection. Struct Health Monit2024; 23: 2270–2287.
27.
WanHPNiYQ.Bayesian multi-task learning methodology for reconstruction of structural health monitoring data. Struct Health Monit2019; 18: 1282–1309.
28.
CorballyRMalekjafarianA.A deep-learning framework for classifying the type, location, and severity of bridge damage using drive-by measurements. Comput Aided Civ Inf2024; 39: 852–871.
29.
LiuBZhangJLeiM, et al. Simultaneous tunnel defects and lining thickness identification based on multi-tasks deep neural network from ground penetrating radar images. Autom Constr2023; 145: 104633.
30.
ShaoWSunHZhouQ, et al. A novel Lamb wave-based multi-damage dataset construction and quantification algorithm under the framework of multi-task deep learning. Struct Health Monit2023; 23: 1148–1169.
31.
LiuJXXuSSBergésM, et al. HierMUD: hierarchical multi-task unsupervised domain adaptation between bridges for drive-by damage diagnosis. Struct Health Monit2023; 22: 1941–1968.
32.
SunTShaoYLiX, et al. Learning sparse sharing architectures for multiple tasks. arXiv:1911.05034.
33.
ZhangYXieXLiH, et al. Subway tunnel damage detection based on in-service train dynamic response, variational mode decomposition, convolutional neural networks and long short-term memory. Autom Constr2022; 139: 104293.
34.
MaJZhaoZYiX, et al. Modeling task relationships in multi-task learning with multi-gate mixture-of-experts. In: Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery and data mining, London, 2018, pp. 1930–1939.
35.
MaJZhaoZChenJ, et al. SNR: sub-network routing for flexible parameter sharing in multi-task learning. In: Proceedings of the thirty-third AAAI conference on artificial intelligence and thirty-first innovative applications of artificial intelligence conference and ninth AAAI symposium on educational advances in artificial intelligence, Honolulu, HI, USA, 2019, Article 27.
36.
TangHLiuJZhaoM, et al. Progressive layered extraction (PLE): a novel multi-task learning (MTL) model for personalized recommendations. In: Proceedings of the 14th ACM conference on recommender systems, Virtual Event, 2020, pp. 269–278.
37.
DosovitskiyABeyerLKolesnikovA, et al. An image is worth 16 × 16 words: transformers for image recognition at scale. arXiv:2010.11929.
38.
VandenhendeSGeorgoulisSGansbekeWV, et al. Multi-task learning for dense prediction tasks: a survey. IEEE Trans Pattern Anal Mach Intell2022; 44: 3614-3633.
