Integrated approach to mine seismicity prediction: Integration of time series and neural networks

The research aims at development of a prediction procedure for geodynamic events in mines. The presented approach to the analysis of seismic data integrates the time series, machine learning and neural network methods. The retrospective analysis of seismic data collected at the Sheregesh deposit between 2016 and 2024 was performed. The patterns were determined using clustering and auto-correlation, which revealed a first-order autoregressive process with a coefficient of 0.422. Clustering allows identifying zones of different seismic activity and anomalous events. The time invariance tests produce conflicting results, which is explained by the seasonal and cyclic nature of mining operations. A nonstationarity is connected with the manmade factors. For predicting average daily energy of events, a recurrent neural network with the long short-term memory was constructed and learned using 90-day windows of seismic data. The model proved effective in case of long-term relationships and is applicable for seismic activity prediction in a variable geomechanical environment. The standard quality metrics are determined: mean absolute error–0.19, mean square error–0.31. This research contributes to the development of methods of geodynamic risk prediction and proposes adaptive solutions for the mining industry.

Konurin A. I., Orlov D. V. Integrated approach to mine seismicity prediction: Integration of time series and neural networks. MIAB. Mining Inf. Anal. Bull. 2025;(10):140-152. [In Russ]. DOI: 10.25018/0236_1493_2025_10_0_140.

The study was supported by the Russian Science Foundation, Grant No. 25-27-00034, https://rscf.ru/project/25-27-00034/.

A.I. Konurin¹, Cand. Sci. (Eng.), Senior Researcher, e-mail: konurin@misd.ru, ORCID ID: 0000-0003-3373-2382,
D.V. Orlov¹, Graduate Student, Engineer, e-mail: dmiorl@gmail.com,
¹ N.A. Chinakal Mining Institute SB RAS, 630091, Novosibirsk, Russia.

A.I. Konurin, e-mail: konurin@misd.ru.

1. Pelipenko M. V., Ainbinder I. I., Rylnikova M. V. Principles of assessing the risk of accidents during the operation of underground mines. News of the Tula state university. Sciences of Earth. 2021, no. 4, pp. 178—192. [In Russ]. DOI: 10.46689/2218-5194-2021-4-1-178-192.

2. Sidorov D. V., Potapchuk M. I., Sidlyar A. V. Forecasting the rock burst hazard of a tectonically disturbed ore massif at deep horizons of the Nikolaev polymetallic deposit. Journal of Mining Institute. 2018, vol. 234, pp. 604—611. [In Russ]. DOI: 10.31897/PMI.2018.6.604.

3. Lomov M. A., Sidlyar A. V. Assessment of rock burst hazard factors of the Nikolaevskoye deposit using a 3D modeling system for seismoacoustic monitoring results. Problems of Subsoil Use. 2021, no. 1, pp. 64—72. [In Russ]. DOI: 10.25635/2313-1586.2021.01.064.

4. Korchak P. A., Karasev M. A. Geomechanical substantiation of the formation of zones of brittle rock failure in the vicinity of the junctions of mine workings of JSC «APATIT». Sustainable Development of Mountain Territories. 2023, vol. 15, no. 1 (55), pp. 67—80. [In Russ]. DOI: 10.21177/1998-4502-2023-15-1-67-80.

5. Kozyrev A. A., Batugin A. S., Zhukova S. A. On the influence of massif water content on its seismic activity during the development of Khibiny apatite deposits. Gornyi Zhurnal. 2021, no. 1, pp. 31—36. [In Russ]. DOI: 10.17580/gzh.2021.01.06.

6. Kozyrev A. A., Zhuravleva O. G., Zhukova S. A. Spatio-temporal variations in seismicity in the area of the Saami fault (Khibiny massif, Kola Peninsula). Gornyi Zhurnal. 2023, no. 1, pp. 79—84. [In Russ]. DOI: 10.17580/gzh.2023.01.13.

7. Grunin A. P., Sidlyar A. V., Kosmatov S. B. Reducing the error of seismoacoustic event location in the PROGNOZ-ADS system of geomechanical monitoring of rock massif. Bulletin of Pacific national university. 2024, no. 1 (72), pp. 13—20. [In Russ].

8. Avdeev A. N., Sosnovskaya E. L. Justification of rational parameters of systems for developing inclined veins of low and medium power under changing cryogenic conditions. News of the Tula state university. Sciences of Earth. 2022, no. 2, pp. 157—168. [In Russ]. DOI: 10.46689/2218-5194-20222-1-157-168.

9. Besedina A. N., Kishkina S. B., Kocharyan G. G. Parameters of sources of a swarm of microseismic events initiated by an explosion at the Korobkovskoye iron ore deposit. Physics of the Earth. Физика Земли. 2021, no. 3, pp. 63—81. [In Russ]. DOI: 10.31857/S0002333721030030.

10. Batugin A. S. Geodynamic effects of the ultimate stress state of the earth's crust. Russian Mining Industry Journal. 2023, no. S1, pp. 14—21. [In Russ]. DOI: 10.30686/1609-9192-2023-S1-14-21.

11. Besedina A. N., Gridin G. A., Kocharyan G. G., Morozova K. G., Pavlov D. V. Activation of seismoacoustic events after mass explosions at the iron ore deposit of the Kursk magnetic anomaly. Journal of Mining Sciences. 2024, no. 1, pp. 3—14. [In Russ]. DOI: 10.15372/FTPRPI20240101.

12. Lovchikov A. V. New concept of the mechanism of rock bursts and other dynamic phenomena for ore deposit conditions. Mining Science and Technology (Russia). 2020, vol. 5, no. 1, pp. 30—38. [In Russ]. DOI: 10.17073/2500-0632-2020-1-30-38.

13. Gvishiani A. D., Panchenko V. Ya., Nikitina I. M. Systems analysis of big data for earth sciences. Herald of the Russian Academy of Sciences. 2023, vol. 93, no. 6, pp. 518—525. [In Russ]. DOI: 10.31857/S0869587323060087.

14. Konurin A., Neverov S., Neverov A., Orlov D., Zharov I., Konurina M. Application of artificial neural networks for stress state analysis based on the photoelastic method. Geohazard Mechanics. 2023, vol. 1, no. 2, pp. 128—139. DOI: 10.1016/j.ghm.2023.03.001.

15. Eremenko A. A., Mulev S. N., Shtirts V. A. Monitoring of geodynamic phenomena by the microseismic method in the development of rock burst hazardous deposits. Journal of Mining Sciences. 2022, no. 1, pp. 12—22. [In Russ]. DOI: 10.15372/FTPRPI20220102.

16. Romanevich K. V., Mulev S. N. Automation of classification of seismic events during seismic monitoring of a coal mine using machine learning. Russian Mining Industry Journal. 2023, no. S5, pp. 58—64. [In Russ]. DOI: 10.30686/1609-9192-2023-5S-58-64.

17. Mousavi S. M., Ellsworth W. L., Zhu W., Chuang L. Y., Beroza G. C. Earthquake transformeran attentive deep-learning model for simultaneous earthquake detection and phase picking. Nature Communications. 2020, vol. 11, no. 1, pp. 1—12. DOI: 10.1038/s41467-020-17591-w.

18. Alsharef A., Aggarwal K., Sonia G., Kumar M., Mishra A. Review of ML and AUTOML solutions to forecast time-series data. Archives of Computational Methods in Engineering. 2022, vol. 29, no. 7, pp. 5297—5311. DOI: 10.1007/s11831-022-09765-0.

19. Sherstinsky A. Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network. Physica D: Nonlinear Phenomena. 2020, vol. 404, article 132306. DOI: 10.1016/j. physd.2019.132306.

20. Sarker I. H. Deep learning: a comprehensive overview on techniques, taxonomy, applications and research directions. SN Computer Science. 2021, vol. 2, no. 6. DOI: 10.1007/s42979-021-00815-1.

21. Di Y., Wang E., Li Z., Liu X., Huang T., Yao J. Comprehensive early warning method of microseismic, acoustic emission, and electromagnetic radiation signals of rock burst based on deep learning. International Journal of Rock Mechanics and Mining Sciences. 2023, vol. 170, article 105519. DOI: 10.1016/j.ijrmms.2023.105519.