| description abstract | The classification and real-time identification of surrounding rock grades are of great significance to guide tunnel boring machine (TBM) construction. Taking a water diversion project in China as the background, this study evaluates the relationship between tunneling parameters, surrounding rock grades, and tunneling performance (TP) using actual TBM tunneling and geological data with the aim to create a novel surrounding rock classification method that is based on tunneling performance. The advance rate and specific energy of excavation are used as the measurement indexes of TBM TP to carry out the surrounding rock classification. The rationality of TP classification is verified by statistical analysis of the distribution of the thrust, torque, rotational speed, penetration, penetration rate, and utilization of surrounding rocks at all TP grades. Furthermore, a cross-validation support vector machine model based on mean interval value backpropagation is used to realize the real-time identification of TP grades, achieving an accuracy of 95%. Finally, the correlation between thrust and penetration rate under the basic quality index (BQ) and TP classification systems is analyzed, respectively. It is found that there is a positive correlation under the TP classification system and a negative correlation under the BQ classification system, indicating that the TP classification system is more suitable for guiding TBM construction than the BQ classification system. The proposed tunneling performance classification method and real-time identification model of TBM tunnel surrounding rocks provide a reference for guiding the safe and efficient tunneling of TBMs. | |
