Abstract:This study considers rock mass as three different classifications: Class II, III, and IV. The initial data set collection included nine parameters: uniaxial compressive strength, Brazilian tensile strength, rock integrity index (Kv), rock wear resistance index (CAI), deformation modulus (E0), Poisson's ratio (μ), net tunneling rate (PR), rotation speed (RPM), and cutterhead thrust (TF). The final sample data set consisted of 150 carefully selected sets of representative engineering measured data. The training data set shows variation in the percentages of different rock mass classes. The percentages of various rock mass classes in the testing data set, however, correspond to the real engineering situation. The percentage distribution of rock mass classes in the sample data set: (a) Training data set and (b) Test data set. Due to the complexity of underground engineering geology, the tunnel boring machine (TBM) usually shows poor adaptability to the surrounding rock mass, leading to machine jamming and geological hazards. For the TBM project of Lanzhou Water Source Construction, this study proposed a neural network called PCA–GRU, which combines principal component analysis (PCA) with gated recurrent unit (GRU) to improve the accuracy of predicting rock mass classification in TBM tunneling. The input variables from the PCA dimension reduction of nine parameters in the sample data set were utilized for establishing the PCA–GRU model. Subsequently, in order to speed up the response time of surrounding rock mass classification predictions, the PCA–GRU model was optimized. Finally, the prediction results obtained by the PCA–GRU model were compared with those of four other models and further examined using random sampling analysis. As indicated by the results, the PCA–GRU model can predict the rock mass classification in TBM tunneling rapidly, requiring about 20 s to run. It performs better than the previous four models in predicting the rock mass classification, with accuracy A, macro precision MP, and macro recall MR being 0.9667, 0.963, and 0.9763, respectively. In Class II, III, and IV rock mass prediction, the PCA–GRU model demonstrates better precision P and recall R owing to the dimension reduction technique. The random sampling analysis indicates that the PCA–GRU model shows stronger generalization, making it more appropriate in situations where the distribution of various rock mass classes and lithologies change in percentage.

Prediction model of rock mass class using classification and regression tree integrated AdaBoost algorithm based on TBM driving data

Real-Time Prediction of TBM Driving Parameters Using Geological and Operation Data<italic/>

Performance Prediction of Hard Rock TBM Based on Extreme Learning Machine

Real-time hard-rock tunnel prediction model for rock mass classification using CatBoost integrated with Sequential Model-Based Optimization

Prediction of geological conditions for a tunnel boring machine using big operational data

Probabilistic machine learning approach to predict incompetent rock masses in TBM construction

Advance Prediction of Rockmass Conditions During TBM Tunnelling Based on Cost-Sensitive Learning under Imbalance Dataset

Real-time Prediction of Rock Mass Classification Based on TBM Operation Big Data and Stacking Technique of Ensemble Learning

Real-Time Prediction of TBM Driving Parameters Using In Situ Geological and Operation Data

A Rock Mass Strength Prediction Method Integrating Wave Velocity and Operational Parameters Based on the Bayesian Optimization Catboost Algorithm

Grouped machine learning methods for predicting rock mass parameters in a tunnel boring machine‐driven tunnel based on fuzzy C‐means clustering

TBM performance estimation using a classification and regression tree (CART) technique

Estimation of the TBM advance rate under hard rock conditions using XGBoost and Bayesian optimization

Prediction of rock mass class ahead of TBM excavation face by ML and DL algorithms with Bayesian TPE optimization and SHAP feature analysis

Perception model of surrounding rock geological conditions based on TBM operational big data and combined unsupervised-supervised learning

Dynamic Prediction of Rock Mass Classification in the Tunnel Construction Process based on Random Forest Algorithm and TBM in situ Operation Parameters

Prediction of rock mass classification in tunnel boring machine tunneling using the principal component analysis (PCA)–gated recurrent unit (GRU) neural network

Predicting the Performance of Tunnel Boring Machines using Big Operational Data

A LightGBM-based strategy to predict tunnel rockmass class from TBM construction data for building control

Cross-project prediction for rock mass using shuffled TBM big dataset and knowledge-based machine learning methods

Classifying the surrounding rock of tunnel face using machine learning