Predicting Slurry Pressure Balance with a Long Short-Term Memory Recurrent Neural Network in Difficult Ground Condition.

Qiang Wang,Xiongyao Xie,Michael A Mooney,Hongjie Yu,Akbar S Namin

doi:10.1155/2021/6678355

Abstract

The safety of tunneling with shield tunnel boring machines largely depends on the tunnel face pressure, which is currently decided by human operators empirically. Face pressure control is vulnerable to human misjudgment and human errors can cause severe consequences, especially in difficult ground conditions. From a practical perspective, it is therefore beneficial to have a model capable of predicting the tunnel face pressure given operation and the changing geology. In this paper, we propose such a model based on deep learning. More specifically, a long short-term memory (LSTM) recurrent neural network is employed for tunnel face pressure prediction. To correlate with PLC data, linear interpolation is employed to transform the borehole geological data into sequential geological data according to the shield machine position. The slurry pressure in the excavation chamber (SPE) is taken as the output in the case study of Nanning Metro, which is confronted with the clogging problem due to the mixed ground of mudstone and round gravel. The LSTM-based SPE prediction model achieved an overall MAPE and RMSE of 3.83% and 10.3 kPa, respectively, in mudstone rich ground conditions. Factors that influence the model, including different kinds and length of input data and comparison with the traditional machine learning-based model, are also discussed.

Highlights

With the growing demand of urban tunneling, mechanized tunneling has become increasingly popular due to its construction efficiency and low ground disturbance [1, 2]
E input of these three models will be (N, time steps × n features) while the input of the long short-term memory (LSTM) model is (N, time steps, n features). e deep feedforward network (DFN) model structure is similar to the LSTM model, whose hypermeters are determined by the numerical experiments. e hyperparameters of the RF and SVR models are obtained via a randomized search and 3-fold cross-validation [39]
A case study of the Nanning Metro Tunnel project is included for model demonstration. e conclusions of the paper are as follows: (1) It is suitable for the LSTM network to deal with big data time series prediction problem due to its ability to take the effect of history inputs into account

Summary

Introduction

With the growing demand of urban tunneling, mechanized tunneling has become increasingly popular due to its construction efficiency and low ground disturbance [1, 2]. Erefore, it would be beneficial to have a model predicting the pressure response and assisting tunneling by suggesting operations in difficult ground conditions. With due respect to the value of a seasoned operator, such practice is not ideal when dealing with the difficult ground (e.g., variable geology, mixed face conditions, high clogging potential, and gas-richness), as the relationship of the slurry pressure between the excavation chamber and the working chamber may be ever-changing [3]. Such a Computational Intelligence and Neuroscience model would be helpful for the automation driving of the shield machine

Methods

Discussion

Conclusion