|
|
|
| Comparison of water inflow prediction methods of hydraulic diversion tunnels during construction |
| ZHOU Caigui1, LI Jing2*, LIANG Qingguo2, CHEN Kelin2
|
| （1. Northwest Water Conservancy & Hydropower Engineering Co., Ltd., Xi'an 710100, Shaanxi, China;2. School of Civil Engineering, Lanzhou Jiaotong University, Lanzhou 730070, Gansu, China） |
|
|
|
|
Abstract Aiming at the water inrush disaster that was easy to occur during tunnel construction, the internal law of tunnel water inrush was analyzed through the statistics cases of tunnel water inrush. The water inflow in tunnels during construction was predicted by using the methods of long short-term memory neural network(LSTM), Elman neural network and multiple linear regression based on partial least square respectively, and compared with the actual water inflow, then the optimal method for predicting the tunnel water inflow was obtained. The results showed that water inrush accidents were more likely to occur in shallow-buried, long tunnels and extra-long tunnels, and in fault, karst and soluble rock strata. By comparing the prediction results of three different models with the water inflow during tunnel construction, the LSTM model had higher accuracy in predicting the water inflow in tunnels during construction.
|
|
Received: 04 October 2022
Published: 20 March 2023
|
|
|
|
|
|
|
|
