Study on the Classification and Identification Methods of Surrounding Rock Excavatability Based on the Rock-Breaking Performance of Tunnel Boring Machines

Rock mass conditions are extremely sensitive to tunnel boring machine (TBM) tunneling. Therefore, establishing a surrounding rock excavatability (SRE) classification system applicable TBM tunnels. Accurately and intelligently identifying grades can also facilitate efficient tunneling intelligent construction. Specific excavation penetration rates were used evaluate SRE. Their correlations with geological parameters explored using the field data from two water conveyance tunnels in China different lithologies. A high-precision empirical SRE was constructed TOPSIS for multi-objective decision-making, it verified engineering cases. An identification model stable phase of cycle established 12,382 rock-breaking datasets deep forest models. Ten characteristic parameters, e.g., total thrust, selected as input features. Hyperparameter optimization achieved grid search method. Deep compared decision tree, random forest, support vector classifier, neural network. The contribution model’s features measured forest. main conclusions follows: proposed method is feasible matches well actual excavation. In classification, accuracy F1 scores when 96.33% 0.9581, respectively. exhibited better grade performance than four Among ten features, most important feature input, while top shield cylinder rod’s chamber pressure least important. findings provide some references prediction control.