Data-driven unsupervised learning models are widely regarded as valuable tools for bridge damage detection. However, variable environmental conditions introduce nonlinear and non-Gaussian characteristics in modal frequencies, along with local spatiotemporal evolutionary properties that can structural damage-induced changes and lead to misdetection. To address these issues, an unsupervised adaptive modal neighborhood standardization (AMNS) method is proposed for damage detection with inherent insensitivity to environmental variabilities. First, several predominant features of modal frequencies are nonlinearly extracted via exponential slow feature analysis to eliminate redundant information. Second, adhering to the concept that similar environmental excitations yield analogous modal responses, the localized modeling subset corresponding to the first nearest neighborhood of each predominant feature is adaptively searched. Subsequently, the localized mean and standard deviation of each predominant feature are employed for equivalent standardization prior to a real-time AMNS model, thereby enabling the model to be insensitive to environmental variations. Third, anomaly-sensitive damage indicators are derived from the Mahalanobis squared distance and paired with real-time dynamic thresholds. Finally, the proposed model is validated on two typical bridge cases, demonstrating its capacity to tolerate nonlinear environmental influences and accurately identify abnormal frequency changes in a timely manner caused by structural damage.
Fernandez-NavamuelAPardoDMagalhãesF, et al. Deep neural network for damage detection in Infante Dom Henrique bridge using multi-sensor data. Struct Health Monit2025; 24(1): 372–401. https://doi.org/10.1177/14759217241227455
2.
AlhamaydehMAswadNG. Structural health monitoring techniques and technologies for large-scale structures: challenges, limitations, and recommendations. Pract Period Struct2022; 27(3): 03122004. https://doi.org/10.1061/(ASCE)SC.1943-5576.0000703
3.
AnYHChatziESimSH, et al. Recent progress and future trends on damage identification methods for bridge structures. Struct Control Health Monit2019; 26(10): e2416. https://doi.org/10.1002/stc.2416
4.
XiaQTianYDCaiDX, et al. Structural flexibility identification and fast Bayesian-based uncertainty quantification of a cable-stayed bridge. Eng Struct2020; 214: 110616. https://doi.org/10.1016/j.engstruct.2020.110616
5.
HwangDYKimSKimHK. Deep Gaussian process regression for damping of a long-span bridge under varying environmental and operational conditions. J Civ Struct Health Monit2023; 13(6): 1431–1445. https://doi.org/10.1007/s13349-023-00710-8
6.
YaoXJYiTHQuCX. Autoregressive spectrum-guided variational mode decomposition for time-varying modal identification under nonstationary conditions. Eng Struct2022; 251: 113543. https://doi.org/10.1016/j.engstruct.2021.113543
7.
PirròMPereiraSGentileC, et al. Damage detection under environmental and operational variability using the cointegration technique. Eng Struct2025; 328: 119700. https://doi.org/10.1016/j.engstruct.2025.119700
SarmadiHEntezamiAMagalhãesF. Unsupervised data normalization for continuous dynamic monitoring by an innovative hybrid feature weighting-selection algorithm and natural nearest neighbor searching. Struct Health Monit2023; 22(6): 4005–4026. https://doi.org/10.1177/14759217231166116
10.
AvciOAbdeljaberOKiranyazS, et al. A review of vibration-based damage detection in civil structures: from traditional methods to machine learning and deep learning applications. Mech Syst Signal Process2021; 147: 107077. https://doi.org/10.1016/j.ymssp.2020.107077
11.
JangJSmythAW. Data-driven models for temperature distribution effects on natural frequencies and thermal prestress modeling. Struct Control Health Monit2022; 27(2): e2489. https://doi.org/10.1002/stc.2489
12.
LarssonCAbdeljaberOBolmsvik, et al. Long-term analysis of the environmental effects on the global dynamic properties of a hybrid timber-concrete building. Eng Struct2022; 268: 114726. https://doi.org/10.1016/j.engstruct.2022.114726
13.
SilvaADFigueiredoEMoldovanI. Damage detection approach for bridges under temperature effects using Gaussian process regression trained with hybrid data. J Bridge Eng2022; 27(11): 04022107. https://doi.org/10.1061/(ASCE)BE.1943-5592.0001949
14.
JinSSChoSJungHJ. Adaptive reference updating for vibration-based structural health monitoring under varying environmental conditions. Comput Struct2015; 158: 211–224. https://doi.org/10.1016/j.compstruc.2015.06.001
15.
ZhouHFNiYQKoJM. Structural damage alarming using auto-associative neural network technique: exploration of environment-tolerant capacity and setup of alarming threshold. Mech Syst Signal Process2011; 25(5): 1508–1526. https://doi.org/10.1016/j.ymssp.2011.01.005
16.
HuangJZLiDSLiHN. A new regime-switching cointegration method for structural health monitoring under changing environmental and operational conditions. Measurement2023; 212: 112682. https://doi.org/10.1016/j.measurement.2023.112682
17.
FigueiredoERaduLWordenK, et al. A Bayesian approach based on a Markov-chain Monte Carlo method for damage detection under unknown sources of variability. Eng Struct2014; 80: 1–10. https://doi.org/10.1016/j.engstruct.2014.08.042
18.
SarmadiHEntezamiASaeedi RazaviB, et al. Ensemble learning-based structural health monitoring by Mahalanobis distance metrics. Struct Control Health Monit2021; 28(2): e2663. https://doi.org/10.1002/stc.2663
19.
BullLAWordenKFuentesR, et al. Outlier ensembles: a robust method for damage detection and unsupervised feature extraction from high-dimensional data. J Sound Vib2019; 453: 126–150. https://doi.org/10.1016/j.jsv.2019.03.025
20.
YanAMKerschenGDe BoeP, et al. Structural damage diagnosis under varying environmental conditions-Part I: a linear analysis. Mech Syst Signal Process2005; 19(4): 847–864. https://doi.org/10.1016/j.ymssp.2004.12.002
21.
ReyndersEWurstenGDe RoeckG. Output-only structural health monitoring in changing environmental conditions by means of nonlinear system identification. Struct Health Monit2014; 13(1): 82–93. https://doi.org/10.1177/1475921713502836
22.
ComanducciGMagalhaesFUbertiniF, et al. On vibration-based damage detection by multivariate statistical techniques: application to a long-span arch bridge. Struct Health Monit2016; 15(5): 515–534. https://doi.org/10.1177/1475921716650630
23.
SarmadiHYuenKV. Early damage detection by an innovative unsupervised learning method based on kernel null space and peak-over-threshold. Comput Aided Civ Inf2021; 36(9): 1150–1167. https://doi.org/10.1111/mice.12635
24.
ZangJGHuangHBSunDS. Subdomain principal component analysis for damage detection of structures subjected to changing environments. Eng Struct2023; 288: 1.1–1.12. https://doi.org/10.1016/j.engstruct.2023.116265
25.
EntezamiASarmadiHBehkamalB. Removal of freezing effects from modal frequencies of civil structures for structural health monitoring. Eng Struct2024; 319: 118722. https://doi.org/10.1016/j.engstruct.2024.118722
26.
WangZYiTHYangDH, et al. Early anomaly warning of environment-induced bridge modal variability through localized principal component differences. ASCE-ASME J Risk Uncertainty Eng Syst Part A Civ Eng2022; 8(4): 04022044. https://doi.org/10.1061/AJRUA6.0001269
27.
WangZYiTHYangDH, et al. Data-driven modal equivalent standardization for early damage detection in bridge structural health monitoring. J Eng Mech2023; 149(1): 06022004. https://doi.org/10.1061/JENMDT.EMENG-6778
28.
SarmadiHKaramodinA. A novel anomaly detection method based on adaptive Mahalanobis-squared distance and one-class kNN rule for structural health monitoring under environmental effects. Mech Syst Signal Process2020; 140: 106495. https://doi.org/10.1016/j.ymssp.2019.106495
29.
YuWKZhaoCH. Recursive exponential slow feature analysis for fine-scale adaptive processes monitoring with comprehensive operation status identification. IEEE Trans Ind Inform2018; 15(6): 3311–3323. https://doi.org/10.1109/TII.2018.2878405
30.
KullaaJ. Distinguishing between sensor fault, structural damage, and environmental or operational effects in structural health monitoring. Mech Syst Signal Process2011; 25(8): 2976–2989. https://doi.org/10.1016/j.ymssp.2011.05.017
31.
PeetersBDe RoeckG. One-year monitoring of the Z24-Bridge: environmental effects versus damage events. Earthq Eng Struct D2001; 30(2): 149–171. https://doi.org/10.3390/s20143999
