SpectralTCN: A multi-resolution wavelet gating network for vibration-based structural damage detection
- 1Faculty of Civil Engineering, University of Transport Technology, Ha Noi, Viet Nam
- 2Application of Artificial Intelligence for Structural Health Monitoring (AI-SHM) Research Group (R2026-UTC-01), Faculty of Civil and Environmental Engineering, University of Transport and Communications, Hanoi, Vietnam
Abstract
Vibration-based structural health monitoring (SHM) detects damage by identifying subtle changes in a structure’s dynamic response. Deep learning models have shown strong potential for automating this classification task. However, most existing architectures share a key limitation: they either process signals purely in the time domain, or apply a global frequency-domain transform that discards information about when each frequency component occurs. This is a serious drawback, because structural damage typically manifests as short, localised events confined to specific frequency bands. This paper proposes SpectralTCN, a temporal convolutional network augmented with a Wavelet Gating Module (WGM) that performs learnable, data-dependent multi-resolution filtering within each convolutional block. The WGM decomposes intermediate features via the Discrete Wavelet Transform (DWT) into physically interpretable sub-bands corresponding to distinct structural vibration modes, applies input-adaptive sigmoid gates independently at each decomposition level, and reconstructs the filtered signal via the Inverse DWT (IDWT) with a learnable residual connection initialised to zero. Unlike FFT-based approaches, the DWT simultaneously preserves both time and frequency information, enabling the network to detect both the timing and the spectral location of damage-induced anomalies. Combined with Generalized Mean (GeM) pooling and large-kernel causal depthwise convolutions, SpectralTCN is evaluated on two benchmark datasets: the Z24 Bridge benchmark and a finite element model (FEM)-derived dataset of the My Thuan cable-stayed bridge. Experiments against 10 baseline models and 4 ablation variants, evaluated via stratified 5-fold cross-validation, demonstrate the effectiveness and generalisation capability of the proposed approach: SpectralTCN attains the highest mean accuracy on both benchmarks (92.2% on Z24 and 92.1% on My Thuan) and outperforms the strongest baseline in the 5-fold cross-validation protocol. In addition, the proposed architecture operates on short, streaming acceleration windows with a purely convolutional backbone of moderate computational cost, making SpectralTCN suitable for near-real-time, online damage detection in continuous bridge monitoring.
Related
- Structural Resolution in Feature Attribution Analysis of Molecular Graph Neural Networks
- Field study on compaction effect and energy evolution characteristics of highway subgrade under low-amplitude vibration
- Bi-objective optimization of simultaneous vibration suppression and energy harvesting in suspension systems based on mixed H2/H∞ control
- Equivalent linearization modeling and experimental validation of natural frequency and vibration isolation performance in metal braided hoses
- Study on vibration characteristics of shearer spiral drum and fatigue life prediction of cutting unit housing under multi-factor coupling
BibTeX
@article{hong2026spectraltcn,
title = {{SpectralTCN: A multi-resolution wavelet gating network for vibration-based structural damage detection}},
author = {Quan Pham Hong and Trung Vu Manh and Bich Nguyen Thach and Le Nguyen Dan and Hoa Tran Ngoc},
journal = {PLOS One},
year = {2026},
month = sep,
doi = {10.1371/journal.pone.0358224},
url = {https://doi.org/10.1371/journal.pone.0358224},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - SpectralTCN: A multi-resolution wavelet gating network for vibration-based structural damage detection
AU - Quan Pham Hong
AU - Trung Vu Manh
AU - Bich Nguyen Thach
AU - Le Nguyen Dan
AU - Hoa Tran Ngoc
PY - 2026
DA - 2026/09/21
JO - PLOS One
DO - 10.1371/journal.pone.0358224
UR - https://doi.org/10.1371/journal.pone.0358224
AB - Vibration-based structural health monitoring (SHM) detects damage by identifying subtle changes in a structure’s dynamic response. Deep learning models have shown strong potential for automating this classification task. However, most existing architectures share a key limitation: they either process signals purely in the time domain, or apply a global frequency-domain transform that discards information about when each frequency component occurs. This is a serious drawback, because structural damage typically manifests as short, localised events confined to specific frequency bands. This paper proposes SpectralTCN, a temporal convolutional network augmented with a Wavelet Gating Module (WGM) that performs learnable, data-dependent multi-resolution filtering within each convolutional block. The WGM decomposes intermediate features via the Discrete Wavelet Transform (DWT) into physically interpretable sub-bands corresponding to distinct structural vibration modes, applies input-adaptive sigmoid gates independently at each decomposition level, and reconstructs the filtered signal via the Inverse DWT (IDWT) with a learnable residual connection initialised to zero. Unlike FFT-based approaches, the DWT simultaneously preserves both time and frequency information, enabling the network to detect both the timing and the spectral location of damage-induced anomalies. Combined with Generalized Mean (GeM) pooling and large-kernel causal depthwise convolutions, SpectralTCN is evaluated on two benchmark datasets: the Z24 Bridge benchmark and a finite element model (FEM)-derived dataset of the My Thuan cable-stayed bridge. Experiments against 10 baseline models and 4 ablation variants, evaluated via stratified 5-fold cross-validation, demonstrate the effectiveness and generalisation capability of the proposed approach: SpectralTCN attains the highest mean accuracy on both benchmarks (92.2% on Z24 and 92.1% on My Thuan) and outperforms the strongest baseline in the 5-fold cross-validation protocol. In addition, the proposed architecture operates on short, streaming acceleration windows with a purely convolutional backbone of moderate computational cost, making SpectralTCN suitable for near-real-time, online damage detection in continuous bridge monitoring.
SN - 1932-6203
ER -