Scholarlink

<p>Intelligent prediction of roof formation pressure in coal mining face is an important part of coal mine automation, which is of great significance for safe and efficient production. In this study, the irregular manifestation of formation pressure and severe roof formation pressure in shallow buried coal seams in Fengjiata coal mine were studied. Utilizing machine learning method, a ground pressure prediction model was developed and tested in the 3205 integrated mining face of Fengjiata Coal Mine. Extensive formation pressure data were collected, and Machine Learning (ML) techniques including Random Forest (RF) and Long Short-Term Memory (LSTM) were studied to improve the accuracy of formation pressure prediction. The massive formation pressure data were denoised using Jupyter Notebook and Python programming language to make the data curves more consistent with the actual situation. However, comparative studies using histograms, box plots, and scatter plots were also conducted to evaluate and demonstrate the performance of the model. Therefore, hyper-parameter tuning was thoroughly studied in order to achieve efficient denoising results in stratigraphic pressure processing. Random Forest (RF) and Long Short-Term Memory (LSTM) were used to predict the denoised formation pressure data. The results show that the Random Forest (RF) model outperforms the Long Short-Term Memory (LSTM) model in predicting stratigraphic pressure variations, exhibiting high prediction and computational efficiency, with a Root Mean Squared Error (RMSE) of 4.76, and taking 5&#xa0;minutes. It advances the development of intelligent early warning systems and opens up possibilities for future research and applications, thus contributing to safe mining practices in coal mines.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Study on Intelligent Early Prediction Method of Ground Pressure in Fully Mechanized Mining Face of Coal Mine Based on Random Forest and LSTM

  • 期刊论文

  • 发表日期：

    2026年2月11日

查看全文
  • Anamor Samuel Kofi,
  • Yong Yuan,
  • Zhongshun Chen,
  • Bo Li,
  • Zhenghan Qin,
  • Djamla Afia Apolline

摘要

Intelligent prediction of roof formation pressure in coal mining face is an important part of coal mine automation, which is of great significance for safe and efficient production. In this study, the irregular manifestation of formation pressure and severe roof formation pressure in shallow buried coal seams in Fengjiata coal mine were studied. Utilizing machine learning method, a ground pressure prediction model was developed and tested in the 3205 integrated mining face of Fengjiata Coal Mine. Extensive formation pressure data were collected, and Machine Learning (ML) techniques including Random Forest (RF) and Long Short-Term Memory (LSTM) were studied to improve the accuracy of formation pressure prediction. The massive formation pressure data were denoised using Jupyter Notebook and Python programming language to make the data curves more consistent with the actual situation. However, comparative studies using histograms, box plots, and scatter plots were also conducted to evaluate and demonstrate the performance of the model. Therefore, hyper-parameter tuning was thoroughly studied in order to achieve efficient denoising results in stratigraphic pressure processing. Random Forest (RF) and Long Short-Term Memory (LSTM) were used to predict the denoised formation pressure data. The results show that the Random Forest (RF) model outperforms the Long Short-Term Memory (LSTM) model in predicting stratigraphic pressure variations, exhibiting high prediction and computational efficiency, with a Root Mean Squared Error (RMSE) of 4.76, and taking 5 minutes. It advances the development of intelligent early warning systems and opens up possibilities for future research and applications, thus contributing to safe mining practices in coal mines.