A Fast Identification Method for Seismic Responses of Bridge Structures by Integrating Digital Signal Features and Deep Learning
Abstract
:1. Introduction
2. Data Signal Feature
2.1. Source of Bridge Earthquake Response Samples
2.2. Sample Sources of Vehicle-Induced Vibration Response of Bridges
3. Feature Extraction Method of Earthquake Response Signal
3.1. Short-Time Fourier Transform (STFT)
3.2. Continuous Wavelet Transform (CWT)
3.3. Mel Frequency Cepstral Coefficient (MFCC)
4. Seismic Response Identification Method Based on Deep Learning
4.1. Identification Method Based on Sequence Classification
4.2. Recognition Method Based on Image Recognition
4.3. Comparison of Earthquake Response Identification Accuracy
5. Conclusions
- (1)
- Under the limit of the number of samples, compared with the LSTM neural network model established with the original vibration signal, the earthquake vibration recognition rate of the LSTM neural network model established based on the signal processing method is significantly improved, which proves that the signal processing method can effectively extract features and optimize model overfitting;
- (2)
- Compared with the temporal classification under the LSTM neural network model, the image recognition and classification method based on the Resnet50 neural network model has better feature extraction capabilities under small samples, mainly due to the significant reduction in the number of sample features;
- (3)
- In the field of real-time monitoring, the image recognition method based on the Resnet50 neural network + continuous wavelet transform can achieve higher recognition accuracy; in large-scale data recognition and acquisition applications, the sequence classification method based on LSTM neural network + Mel cepstral coefficients has higher cost performance.
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Zhou, K.; Li, Q.-S.; Han, X.-L. Modal Identification of Civil Structures via Stochastic Subspace Algorithm with Monte Carlo–Based Stabilization Diagram. J. Struct. Eng. 2022, 48, 04022066. [Google Scholar] [CrossRef]
- Xu, Y.J.; Lu, X.Z.; Tian, Y.; Huang, Y. Real-time seismic damage prediction and comparison of various ground motion intensity measures based on machine learning. J. Earthq. Eng. 2020, 26, 4259–4279. [Google Scholar] [CrossRef]
- Xu, Y.J.; Lu, X.Z.; Cetiner, B.; Taciroglu, E. Real-time regional seismic damage assessment framework based on long short-term memory neural network. Comput.-Aided Civ. Infrastruct. Eng. 2020, 36, 504–521. [Google Scholar] [CrossRef]
- Lu, X.Z.; Liao, W.J.; Huang, W.; Xu, Y.; Chen, X. An improved linear quadratic regulator control method through convolutional neural network-based vibration identification. J. Vib. Control 2021, 21, 839–853. [Google Scholar] [CrossRef]
- Cardoni, A.; Elahi, A.R.; Cimellaro, G.P. A Refined Output-Only Modal Identification Technique for Structural Health Monitoring of Civil Infrastructures. Eng. Struct. 2025, 323, 119210. [Google Scholar] [CrossRef]
- Stoura, C.D.; Paraskevopoulos, E.; Dimitrakopoulos, E.G.; Natsiavas, S. A Dynamic Partitioning Method to Solve the Vehicle-Bridge Interaction Problem. Comput. Struct. 2021, 251, 106547. [Google Scholar] [CrossRef]
- Yang, X.-M.; Yi, T.-H.; Qu, C.-X.; Li, H.-N.; Liu, H. Modal Identification of High-Speed Railway Bridges through Free-Vibration Detection. J. Eng. Mech. 2020, 146, 04020107. [Google Scholar] [CrossRef]
- Lécun, Y.; Bottou, L.; Bengio, Y.; Haffner, P. Gradient-based learning applied to document recognition. Proc. IEEE 1998, 86, 2278–2324. [Google Scholar] [CrossRef]
- Srivastava, N.; Mansimov, E.; Salakhutdinov, R. Unsupervised learning of video representations using LSTMs. In Proceedings of the 32nd International Conference on Machine Learning, Lille, France, 7–9 July 2015; pp. 843–852. [Google Scholar]
- Krizhevsky, A.; Sutskever, I.; Hinton, G.E. ImageNet classification with deep convolutional neural networks. In Proceedings of the 2012 Advances in Neural Information Processing Systems, Lake Tahoe, NV, USA, 3–6 December 2012; Curran Associates Inc.: New York, NY, USA, 2012; pp. 1097–1105. [Google Scholar]
- Antipov, G.; Berrani, S.A.; Dugelay, J.L. Minimalistic CNN-based ensemble model for gender prediction from face images. Pattern Recognit. Lett. 2016, 70, 59–65. [Google Scholar] [CrossRef]
- Center for Engineering Strong Motion Data [DB/OL]. Available online: https://www.cesmd.org/index.html (accessed on 6 January 2025).
- Cooley, J.W.; Tukey, J.W. An Algorithm for the Machine Calculation of Complex Fourier Series. Math. Comput. 1965, 19, 297–301. [Google Scholar] [CrossRef]
- Hossan, M.A.; Memon, S.; Gregory, M.A. A novel approach for MFCC feature extraction. In Proceedings of the 4th International Conference on Signal Processing and Communication Systems, Gold Coast, QLD, Australia, 13–15 December 2010; IEEE Press: New York, NY, USA, 2010; pp. 1–5. [Google Scholar]
- Nakisa, B.; Rastgoo, M.N.; Rakotonirainy, A.; Maire, F.; Chandran, V. Long Short Term Memory Hyperparameter Optimization for a Neural Network Based Emotion Recognition Framework. IEEE Access 2018, 6, 49325–49338. [Google Scholar] [CrossRef]
Neural Network | Signal Processing Method | Sample Specifications | Test Time for 1000 Samples (s) | 1000 Samples Consume Memory (MB) | Test Set Accuracy |
---|---|---|---|---|---|
NLSTM | Original data | double | 0.92 | 10.94 | 83.20% |
Continuous wavelet transform | double | 5.25 | 1625.00 | 90.60% | |
Short-time Fourier transform | complex double | 0.92 | 21.00 | 93.50% | |
Mel cepstral coefficient | double | 0.90 | 5.24 | 96.10% | |
Resnet50 | Continuous wavelet transform | pixel | 5.80 | 7.35 | 98.80% |
Short-time Fourier transform | pixel | 5.80 | 7.38 | 96.70% | |
Mel cepstral coefficient | pixel | 5.82 | 7.37 | 96.30% |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Lv, Z.; Ding, Y.; Guo, J. A Fast Identification Method for Seismic Responses of Bridge Structures by Integrating Digital Signal Features and Deep Learning. Sensors 2025, 25, 399. https://doi.org/10.3390/s25020399
Lv Z, Ding Y, Guo J. A Fast Identification Method for Seismic Responses of Bridge Structures by Integrating Digital Signal Features and Deep Learning. Sensors. 2025; 25(2):399. https://doi.org/10.3390/s25020399
Chicago/Turabian StyleLv, Zhaoxu, Youliang Ding, and Junxiao Guo. 2025. "A Fast Identification Method for Seismic Responses of Bridge Structures by Integrating Digital Signal Features and Deep Learning" Sensors 25, no. 2: 399. https://doi.org/10.3390/s25020399
APA StyleLv, Z., Ding, Y., & Guo, J. (2025). A Fast Identification Method for Seismic Responses of Bridge Structures by Integrating Digital Signal Features and Deep Learning. Sensors, 25(2), 399. https://doi.org/10.3390/s25020399