Search

In order to solve the problem of mine water surge prediction, we introduce deep learning theory, combine long short-term memory network (LSTM) and gated circulation unit (GRU), select mine water surge as the research object, and establish a mine water surge prediction model based on LSTM-GRU. Taking the mine water inflow of a mine in Shaanxi Province as sample data, the data set was divided into a training set and a test set using a 7∶3 ratio, and the gradient descent algorithm with good model training effect was selected to determine the network model parameters and regularization parameters. In order to prove the prediction accuracy of the LSTM-GRU model, the prediction results were compared with those obtained by the traditional ARIMA model and the LSTM model to predict mine water gusher, respectively. The results show that: the mean absolute percentage error (RMSE), root mean square error (MAE), mean absolute error (MAPE) and coefficient of determination (R2) of the LSTM-GRU composite model are 70.51, 53.4, 2.80% and 0.86, indicating that the model has high prediction accuracy and reliability. The prediction effect is better than the traditional ARIMA model and LSTM model.

The prediction of mine water inflow is of great significance to the safe production of coal mines. The accurate selection of hydrogeological parameters and the selection of calculation methods have a great influence on the accurate prediction of mine water inflow. At the same time, taking into account the actual mining of the mine, the water inflow is not stable from the early stage of mining, but dynamic. In order to improve the accuracy of water inflow prediction and better serve coal mining, based on the analysis of geological and hydrogeological conditions of Donggou Coal Mine in Xinjiang and the selection of accurate hydrogeological parameters, two water inflow prediction methods suitable for Donggou Coal Mine were selected for comparative analysis. The calculation results show that : ① The water inflow calculated by the unsteady flow water release-section flow method shows dynamic change characteristics. After the mining intensity is stable, the change range becomes smaller and finally tends to be stable. ② The predicted water inflow calculated by the large well method and the unsteady flow water release-section flow method is 3396 m3/d, 599.87 − 612.9 m3/d and the actual water inflow 480 m3/d, respectively. The difference with the actual value is in ascending order of magnitude 607.5% > 24.97% − 27.69% ; ③ According to the verification and analysis of the water inflow data generated by the actual mining of the mine, the water inflow calculated by the unsteady flow water release-section flow method is basically consistent with the water inflow generated by the actual mining. The accuracy of the water inflow prediction results is in the order of unsteady flow water release-section flow method > large well method ; ④ In the case of sufficient data and accurate parameters, it is recommended to give priority to the unsteady flow release-section flow method to calculate the water inflow. It is recommended to use the large well method first under the condition that the water supply boundary conditions are poor and meet the requirements of the large well method. On the one hand, the calculation results fully verify the accuracy of the unsteady flow release-section flow method. On the other hand, it provides a reference for the calculation of water inflow in coal mines, and also provides data support for the safe production of Donggou Coal Mine.

Copyright of Coal Science & Technology (0253-2336) is the property of Coal Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)