Earthquake precursor detection algorithm based on two signal decomposition methods and machine learning and its numerical study

Cover Page

Cite item

Full Text

Open Access Open Access
Restricted Access Access granted
Restricted Access Subscription or Fee Access

Abstract

The results of the combined application of the empirical mode decomposition methods, internal decomposition over the time scale and the Hilbert transform for individual modes in order to identify the diagnostic feature of the main event precursor are presented. А computational experiment aimed at a comparative study of the reliability and stability of the obtained forecast for detecting earthquake precursors on a specific sample of real observations against the neural network algorithm was conducted.

Full Text

Restricted Access

About the authors

S. I. Kolesnikova

Saint-Petersburg State University of Aerospace Instrumentation

Author for correspondence.
Email: ksi@guap.ru

Dr. of Tech. Sc., Professor

Russian Federation, St. Petersburg

E. A. Tsygankova

Saint-Petersburg State University of Aerospace Instrumentation

Email: katetsugankova@yandex.ru

Master’s Student

Russian Federation, St. Petersburg

References

  1. Huang N. E., Shen Z., Long S. R., Wu M. C., Shih H. H., Zheng Q., Yen N., Tung C. C., Liu Н. H. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis, Proc. R. SOC. London, Ser. А, 1998, pp. 903—995.
  2. Frei M., Osorio G. I ntrinsic time-scale decomposition: time-frequency-energy analysis and real-time filtering of non-stationary signals, Proc. R. Soc. Lond. А: Mathematical, Physical and Engineering Sciences,2006,vol. 463, pp. 321—342.
  3. Hakim D. K., Gernowo R., Nirwansyah А. W. Flood prediction with time series data mining: Systematic review, Natural Hazards Research, 2024, vol. 4, no. 2, pp. 194—220, doi: 10.1016/j.nhres.2023.10.001.
  4. Çoban B., Scaparra M. P., O’Hanley J. R. Use of OR in earthquake operations management: А review of the literature and roadmap for future research, International Journal of Disaster Risk Reduction, 2021, vol. 65, p. 102539, doi: 10.1016/j.ijdrr.2021.102539.
  5. Abdalzaher M. S., Krichen M., Falcone F. Emerging technologies and supporting tools for earthquake disaster management: А perspective, challenges, and future directions, Progress in Disaster Science, 2024, vol. 23, p. 100347, doi: 10.1016/j.pdisas.2024.100347.
  6. He X., Chen Zh., Yang Q., Xu Ch. Advances in earthquake and cascading disasters, Natural Hazards Research, 2025, in press, doi: 10.1016/j.nhres.2025.01.010.
  7. Pwavodi J., Ibrahim A. U., Pwavodi P. C., Al-Turjman F., Mohand-Said А. The role of artificial intelligence and IoT in prediction of earthquakes: Review, Artificial Intelligence in Geosciences, 2024, vol. 5, p. 100075, doi: 10.1016/j.aiig.2024.100075.
  8. Laurenti L., Tinti E., Galasso F., Franco L., Marone C. Deep learning for laboratory earthquake prediction and autoregressive forecasting of fault zone stress, Earth and Planetary Science Letters, 2022, vol. 598, p. 117825, doi: 10.1016/j.epsl.2022.117825.
  9. Kopylova G. N., Lyubushin A. A., Taranova L. N. New prognostic technology for analyzing variations in low-frequency seismic noise (using the example of regions of the Russian Far East), Rossiyskiy seysmologicheskiy zhurnal, 2021, vol. 3, no. 1, pp. 75—91 (in Russian), doi: 10.35540/2686-7907.2021.1.05.
  10. Bespalko A. A., Yavorovich L. V., Kolesnikova S. I., Bukreev V. G., Mertvetsov A. N., Fedotov P. I. Study of changes in the characteristics of electromagnetic signals during uniaxial compression of rock samples from the Tashtagol mine, Izvestiya vuzov. Fizika, 2011, vol. 54, no. 1—2, pp. 78—85 (in Russian).
  11. Baktibayev D., Baigozha B., Akhmetov I., Mussabayev R., KrassovitskiyA., Toleu А. Literature review on aftershock and earthquake prediction models aided by NLP summarization and ontology extraction techniques, Procedia Computer Science, 2024, vol. 238, no. 2, pp. 579—586, doi: 10.1016/j.procs.2024.06.064.
  12. Mandrikova O. V., Solovjev I. S., Khomutov S. Yu., Baishev D. G., Geppener V. V., Klionskiy D. M. Estimation of Geomagnetic Field Storminess Using the Wavelet Transform, Pattern recognition and image analysis (advances in mathematical theory and applications), 2016, vol. 26, no. 4, pp. 71—79.
  13. Lyu Z., Ororbia A., Desell T. Online evolutionary neural architecture search for multivariate non-stationary time series forecasting, Applied Soft Computing, 2023, vol. 145, 110522, doi: 10.1016/j.asoc.2023.110522.
  14. Kumar A., Wyłomańska A., Zimroz R., Xiang J., Antoni J. Critical challenges and advances in vibration signal processing for non-stationary condition monitoring, Advanced Engineering Informatics, 2025, vol. 65, part C, p. 103290, doi: 10.1016/j.aei.2025.103290.
  15. Poulimenos A. G., Fassois S. D. Parametric time-domain methods for non-stationary random vibration modelling and analysis — А critical survey and comparison, Mechanical Systems and Signal Processing, 2006, vol.20, iss. 4, pp. 763—816, doi: 10.1016/j.ymssp.2005.10.003.
  16. Voznesensky А. S. Adaptive numerical methods of filtering and spectral analysis of non-stationary signals based on time-frequency decomposition: dis. ... Cand. of Engineering Sciences: defended 06/09/2022: approved 04/08/2022. St. Petersburg, 141 p. (in Russian).
  17. Panov S. V., Parushkin M. D., Semibalamut V. M., Fomin Yu. N. Application of empirical mode decomposition for monitoring deformation processes in adits and workings, Fiziko-tekhnicheskiye problemy razrabotki poleznykh iskopayemykh, 2017, no. 5, pp. 186—194 (in Russian).
  18. Panov S. V., Parushkin M. D., Semibalamut V. M., Fomin Yu. N. Heterodyne laser strainmeter for geophysical research, Aktual’nyye problemy sovremennoy seysmologii, Tashkent, 2016, pp. 293—299 (in Russian).
  19. Dyuk V. A., Komashinsky V. I., Malygin I. G., Senkevich Yu. I., Avanesov M. Yu. Study of empirical mode decomposition in the problem of analyzing acoustic emission signals, Informatsiya i kosmos, 2018, no. 4, pp. 50—55 (in Russian).
  20. Oparin V. N. Methods and systems of seismic deformation monitoring of man-made earthquakes and rock bursts: in 2 vol’s, vol. 2 / ed. N. N. Melnikov, Novosibirsk, Publishing house of SO RAN, 2010, 261 p. (in Russian).
  21. Davis J., Goadrich M. The relationship between Precision-Recall and ROC curves, Proceedings of the 23rd International Conference on Machine Learning — ICML, 2006, pp. 233—240, doi: 10.1145/1143844.1143874.
  22. Perol T., Gharbi M., Denolle M. Convolutional neural network for earthquake detection and location, Science Advances, 2018, vol. 4 (2), p. 1700578, doi: 10.1126/sciadv.1700578.
  23. Sugiyama D., Tsuboi S., Yukutake Y. Application of deep learning-based neural networks using theoretical seismograms as training data for locating earthquakes in the Hakone volcanic region, Japan, Earth Planets Space, 2021, vol. 73 (135), doi: 10.1186/s40623-021-01461-w.
  24. Ulyanov N. A., Yaskevich S. V., Dergach P. A., Yablokov А. V. Detection of records of weak local earthquakes using neural networks, Geofizicheskiye tekhnologii, 2021, no. 2, pp. 13—23 (in Russian).

Supplementary files

Supplementary Files
Action
1. JATS XML
Download
2. Fig. 1. Result of decomposition of the signal Y(t) (Signal 9 from dataset [17]) according to the EMD (a) and ITD (b) algorithms, respectively

Download (242KB)
Indexing metadata
3. Fig. 2. Optimal (during training) mode function under the conditions of the given dataset and illustration of the diagnostic feature obtained on its basis: a – graph of the eighth function (IMF8) based on EMD; b – result of the HHT transform of the eighth function (IMF8) (Signal 10 from dataset [17])

Download (130KB)
Indexing metadata
4. Fig. 3. Fifth function of the ITD decomposition (Rotation 5): a – fragment of the fifth high-frequency function based on ITD; b – result of the HHT transform of Rotation 5

Download (126KB)
Indexing metadata
5. Fig. 4. Dependence of quality metric values for the p-class and n-class on NSR based on AEMD

Download (84KB)
Indexing metadata
6. Fig. 5. Dependence of quality metric values for the p-class and n-class on NSR based on AITD

Download (90KB)
Indexing metadata
7. Fig. 6. Dependence of the Type I error rate on NSR according to the EMD (a) and ITD (b) methods

Download (52KB)
Indexing metadata
8. Fig. 7. Illustration of an example of performing signal analysis functions in the software: signal decomposition, spectral analysis, and application of the decision rule

Download (440KB)
Indexing metadata
9. Fig. 8. Architecture of the hybrid neural network: layers and dimensions of the input data

Download (58KB)
Indexing metadata

Copyright (c) 2026 Informacionnye Tehnologii


СМИ зарегистрировано Федеральной службой по надзору в сфере связи, информационных технологий и массовых коммуникаций (Роскомнадзор).
Регистрационный номер и дата принятия решения о регистрации СМИ: серия ПИ № 77 - 15565 от 02 июня 2003 г.