An open access journal
The prediction of deep coal mining based on grey prediction
Abstract
With increasing coal mining depth, the likelihood of rock bursts has significantly risen, posing a major threat to coal mine safety in China. This paper aims to develop classification and prediction models to identify and predict rock burst precursor signals, thereby mitigating this hazard in deep mining. Using acoustic emission (AE) and electromagnetic radiation (EMR) data, we extracted time-frequency domain features and constructed decision tree and grey prediction models with a sliding window approach. For interference signal identification, we analyzed class C signals, determining their mean, kurtosis, and spectral peak values. For predicting class B signal trends, similar features were used to construct a decision tree model, which successfully identified precursor signals. The model demonstrated excellent classification performance in ROC curve tests. This research provides a scientific basis for preventing and controlling rock bursts, reducing engineering efforts, and enhancing coal mine safety.
Share and Cite
Article Metrics
References
- Chen Xingang, Yang Dingkun, Tan Hao, etc. Analysis of Raman spectroscopy of the characteristic gas in transformer oil based on kurtosis [J]. High-voltage technology, 2017,43(07):2256-2262.DOI:10.13336/j.1003-6520.hve.20170628022.
- Liu Xiaohong, Gong Ruichun, Tong Xiaomei. Research on the approximate analysis of continuous signal frequency spectrum by using discrete Fourier transform [J]. Computer CD software and application, 2014,17 (21): 137-138.
- Tian Teng, Shi Maolin, Song Xuegong, et al. Time-series anomaly detection method based on sliding windows [J]. Instrument Technology and Sensors, 2021, (07): 112-116.
- CAI Weiling, Chen Dongxia. The effect of the data normalization method on the K-nearest neighbor classifier [J]. Computer Engineering, 2010,36 (22): 175-177.
- Yang Xuebing, Zhang Jun. The decision tree algorithm and its core techniques [J]. Computer Technology and Development, 2007, (01): 43-45.
- Luo Jia, Li Mingming. Design and application of student employment prediction model based on decision tree algorithm [J]. Integrated circuit applications, and 2023,40(10):62-64.DOI:10.19339/j.issn.1674-2583.2023.10.024.
- Wang Qinghe, Wang Qingshan. Several commonly used digital filtering algorithms in data processing [J]. Metering Technology, 2003, (04): 53-54.
- Xie Jiecheng, Zhang Dali, Xu Wenli. Summary of wavelet image denoising [J]. Chinese Journal of Image and graphics, 2002 (03): 3-11.
- Zhao Ziyi, Hao Zhongqi, Lu Ying, et al. Improvement of the stability of laser-induced breakdown spectrum detection by two-dimensional wavelet denoising [J / OL]. Journal of Optics, 1-14 [2024-05-03].
- Zhang Jun. Improvement of the grey prediction model and its application [D]. Xi'an University of Technology, 2008.
- Yang Hualong, Liu Jinxia, Zheng Bin. Improvement and application of grey prediction GM (1,1) model [J]. The Practice and Awareness of Mathematics, 2011,41 (23): 39-46
- Zheng Jian, Chen Bin. GM (1,1) based on time weight series [J]. Control and decision-making, 2018,33(03):529-534.DOI:10.13195/j.kzyjc.2017.0033.