Mine Water Inflow Prediction Using a CEEMDAN-OVMD-Transformer Model
Abstract
1. Introduction
2. Principles of the Prediction Model Based on Empirical Mode Decomposition and Transformer
2.1. Principle of Complete Ensemble Empirical Mode Decomposition with Adaptive Noise
2.2. Optimal Variational Mode Decomposition
2.3. Principle of the Transformer Model
2.4. CEEMDAN-OVMD-Transformer Modeling Principles
3. Data Processing and Model Evaluation Metrics
3.1. Data Temporalization
3.2. Evaluation Indicators
4. Mine Water Inflow Prediction Study Results
4.1. Overview of the Study Area
4.2. Characteristics and Preprocessing of Water Inflow Data
4.3. Decomposition of Mine Water Inflow Data
4.4. Forecast Results and Analysis
5. Discussion
6. Conclusions
- (1)
- By combining CEEMDAN-OVMD decomposition, the original sequence is broken into IMFs with different frequencies and amplitudes, enabling multi-scale separation of mine water inflow data. This converts chaotic mixed data into organized subsequences, helping the Transformer model learn the underlying patterns of water inflow dynamics.
- (2)
- High-frequency IMFs obtained from CEEMDAN undergo secondary decomposition through OVMD to refine the high-frequency signals. This purification decreases the difficulty for the Transformer to process high-frequency components, allowing it to focus on learning true patterns. As a result, the accuracy of water inflow predictions improves.
- (3)
- Through model comparison, the CEEMDAN-OVMD-Transformer model proposed in this paper employs the CEEMDAN primary decomposition and OVMD secondary refinement within a dual-modal framework to extract complex features from water inflow sequences. By integrating the Transformer to capture long-term temporal dependencies, it thoroughly explores and learns the hidden variation patterns in historical mine water inflow data, leading to improved predictive accuracy for mine water inflow. This method can be applied to short-term water inflow forecasting in similar mines, providing a foundation for mine water prevention and control.
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Hou, E.; Xu, L.; Rong, T. Time series prediction of mine water inflow from Binchang Dafosi mine. J. Xi’an Univ. Sci. Technol. 2024, 44, 490–500. [Google Scholar]
- Yang, J.; Pu, R. Research on the distribution characteristics and water filling modes of mine water inflow in western mining areas of China. Coal Sci. Technol. 2025, 1–13. [Google Scholar]
- Yingying, D.; Shangxian, Y.; Huiqing, L.; Wei, L.; Qixing, L.; Rongrong, Q.; Changsen, B.; Xiangxue, X.; Shuqian, L. Prediction of mine water inflow along mining faces using the SSA-CG-Attention multifactor model. Coal Geol. Explor. 2024, 52, 111–119. [Google Scholar]
- Liu, H.; Liu, G.; Ning, D.; Fan, J.; Chen, W. Mine water inrush prediction method based on VMD-DBN model. Coal Geol. Explor. 2023, 51, 13–21. [Google Scholar]
- Zhang, Z.; Li, W. Comparative Study on Prediction Methods of Mine Inflow: A Case Study of Dazhuang Coal Mine in Gansu Province. Northwestern Geol. 2022, 55, 355–360. [Google Scholar]
- Zhou, Q.; Wang, D.; Dang, Z.; Huo, C.; Su, J.; Liu, C.; Gu, L.; Li, J. Research and comparative analysis of mine water inflow prediction method: A case study of Donggou Coal Mine in Xinjiang. Coal Sci. Technol. 2024, 52 (Suppl. S1), 211–220. [Google Scholar]
- Zhao, Y.; Li, X. Bi-RNN and GMS coupled prediction of water inflow in metal open-pit mines. Coal Geol. Explor. 2024, 52, 155–169. [Google Scholar]
- Li, J.; Gao, P.; Wang, X.; Zhao, S. Prediction of mine water inflow based on Chaos—Generalized Regression Neural Network. Coal Sci. Technol. 2022, 50, 149–155. [Google Scholar]
- Meng, X.; Shi, H. A Survey of Transformer-based Model for Time Series Forecasting. J. Front. Comput. Sci. Technol. 2025, 19, 45–64. [Google Scholar]
- Liang, H.; Liu, S.; Du, J. Review of Deep Learning Applied to Time Series Prediction. J. Front. Comput. Sci. Technol. 2023, 17, 1285–1300. [Google Scholar]
- Liu, X.; Ji, Y.; Zhu, K.; Zhao, C.; Li, K.; Li, C.; Yuan, C.; Li, P.; Yan, P. Construction and application of a TCN-LSTM-SVM-based time series prediction model for water inflow in coal seam roofs. Coal Geol. Explor. 2025, 53, 201–211. [Google Scholar]
- Thirunavukarasu, A.J.; Ting, D.S.J.; Elangovan, K.; Gutierrez, L.; Tan, T.F.; Ting, D.S.W. Large language models in medicine. Nat. Med. 2023, 29, 1930–1940. [Google Scholar] [CrossRef] [PubMed]
- Annan, R.; Qingge, L. Artificial intelligence in COVID-19 research: A comprehensive survey of innovations, challenges, and future directions. Comput. Sci. Rev. 2025, 57, 100751. [Google Scholar] [CrossRef]
- Zhou, G.; Qu, Z.; Zhan, Z. Fusion of wavelet denoising and VMD-NTEO for transmission line fault localization. Proc. CSU-EPSA 2025, 37, 141–150. [Google Scholar]
- Wang, F.; Rong, T.; Hou, E.; Fan, Z.; Tan, E. VMD-BiLSTM Combined Model-based Mine Research on time series prediction method of water influx. Min. Res. Dev. 2024, 44, 143–151. [Google Scholar]
- Qi, X.; Shao, L.; Xing, Y. ARMA prediction model of mine water inflow based on EMD. J. Liaoning Tech. Univ. (Nat. Sci.) 2020, 39, 509–513. [Google Scholar]
- Hou, E.; Xia, B.; Wu, Z.; Rong, T. Mine Water Inflow Prediction Based on CEEMDAN-BO-BiGRU. Sci. Technol. Eng. 2023, 23, 12012–12019. [Google Scholar]
- Torres, M.E.; Colominas, M.A.; Schlotthauer, G.; Flandrin, P. A complete ensemble empirical mode decomposition with adaptive noise. In Proceedings of the IEEE International Conference on Acoustics, Speech Signal Processing, Prague, Czech Republic, 22–27 May 2011; pp. 4144–4147. [Google Scholar]
- Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A.N.; Kaiser, Ł.; Polosukhin, I. Attention Is All You Need. In Proceedings of the Advances in Neural Information Processing Systems, Long Beach, CA, USA, 4–9 December 2017; pp. 3778–3786. [Google Scholar]
- Bian, J.; Hou, T.; Ren, D.; Lin, C.; Qiao, X.; Ma, X.; Ma, J.; Wang, Y.; Wang, J.; Liang, X. Predicting mine water inflow volumes using a decomposition-optimization algorithm-machine learning approach. Sci. Rep. 2024, 14, 17777. [Google Scholar] [CrossRef] [PubMed]
- Qin, J.; Xu, S.; Zhang, B.; Zong, J.; Zhang, J. Research on Prediction Method of Mine Water Inflow Based on CNN-LSTM Model. Coal Technol. 2025, 44, 234–239. [Google Scholar]
- Hu, X.; Li, B.; Shi, Y.; Yin, J.; Li, C.; Yu, Z. NRBO dual-optimized VMD-Transformer model landslide prediction method. Bull. Surv. Mapp. 2025, 1, 143–150. [Google Scholar]
- Yang, J.; Shu, L.; Zhang, S.; Quin, K.; Cui, C. Research on a PCA-transformer-based prediction algorithm for gas concentration in working face. J. Mine Autom. 2025, 51, 1–7. [Google Scholar]
Input Sample | Sample Output |
---|---|
Component | Sample Entropy Value | Frequency Characteristics |
---|---|---|
IMF1 | 1.3418 | High Frequency |
IMF2 | 0.3765 | Low Frequency |
IMF3 | 0.4972 | Low Frequency |
IMF4 | 0.3417 | Low Frequency |
IMF5 | 0.1372 | Low Frequency |
Res | 0.0497 | Low Frequency |
Mode Number | IMF1 | IMF2 | IMF3 | IMF4 | IMF5 |
---|---|---|---|---|---|
K = 2 | 0.2652 | 0.4693 | - | - | - |
K = 3 | 0.1663 | 0.2573 | 0.3197 | - | - |
K = 4 | 0.1674 | 0.2574 | 0.3040 | 0.3687 | - |
K = 5 | 0.1619 | 0.2446 | 0.2823 | 0.3119 | 0.4193 |
Method | MAE (m3/h) | RMSE (m3/h) | MAPE | R2 |
---|---|---|---|---|
CEEMDAN-OVMD-BP | 0.859 | 1.079 | 0.016 | 0.912 |
CEEMDAN-OVMD-SVM | 0.989 | 1.109 | 0.019 | 0.898 |
CEEMDAN-OVMD-LSTM | 0.554 | 0.660 | 0.010 | 0.943 |
CEEMDAN-OVMD-GRU | 0.545 | 0.650 | 0.010 | 0.944 |
Transformer | 0.774 | 0.964 | 0.015 | 0.920 |
CEEMDAN-OVMD-Transformer | 0.507 | 0.613 | 0.010 | 0.948 |
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
Li, Y.; Wu, Q.; Lei, F. Mine Water Inflow Prediction Using a CEEMDAN-OVMD-Transformer Model. Appl. Sci. 2025, 15, 9710. https://doi.org/10.3390/app15179710
Li Y, Wu Q, Lei F. Mine Water Inflow Prediction Using a CEEMDAN-OVMD-Transformer Model. Applied Sciences. 2025; 15(17):9710. https://doi.org/10.3390/app15179710
Chicago/Turabian StyleLi, Yang, Qiang Wu, and Fangchao Lei. 2025. "Mine Water Inflow Prediction Using a CEEMDAN-OVMD-Transformer Model" Applied Sciences 15, no. 17: 9710. https://doi.org/10.3390/app15179710
APA StyleLi, Y., Wu, Q., & Lei, F. (2025). Mine Water Inflow Prediction Using a CEEMDAN-OVMD-Transformer Model. Applied Sciences, 15(17), 9710. https://doi.org/10.3390/app15179710