Intelligent Optimization Method of Shield Tunneling Parameters Based on Machine Learning

Abstract

Aiming at the problems that the parameter setting in the process of shield tunneling depends on artificial experience and is difficult to adapt to complex and changeable geological conditions, this paper studies the intelligent optimization of shield tunneling parameters based on machine learning. Combined with the nonlinear and multi-factor coupling characteristics of shield construction, three machine learning algorithms, random forest (RF), support vector machine (SVM) and BP neural network, are selected to construct a parameter optimization model with geological parameters and tunneling parameters as input and tunneling speed as output. The mean absolute error (MAE), root mean square error (RMSE) and coefficient of determination (R²) were selected as evaluation indexes to evaluate the performance of the model. The results show that all three models can predict the tunneling speed, but the performance difference is significant. Among them, the random forest model performs best in nonlinear mapping and anti-interference ability through ensemble learning and feature random selection. Its R² reaches 0.967, and the prediction accuracy is significantly better than that of BP neural network and support vector machine model. This study provides an effective method support for intelligent prediction and dynamic optimization of shield tunneling parameters, and has practical reference value for improving the intelligent level of shield construction.