Classification and prediction of rock mass drillability for a tunnel boring machine based on operational data mining DOI Creative Commons

Mingshe Sun,

Song Chen,

Huafei He

и другие.

Frontiers in Earth Science, Год журнала: 2024, Номер 12

Опубликована: Дек. 23, 2024

Currently, the accurate prediction of tunnel boring machine (TBM) performance remains a considerable challenge due to complex interactions between TBM and rock mass. In this study, research work is based on part metro project that covers 2,083.94 m. The Gaussian mixture model (GMM) K-nearest neighbor algorithm (KNN) are used classify predict mass drillability in excavation process. Drillability indexes introduced cluster mass, including penetration (P), field index (FPI), torque (TPI), specific energy (SE). Statistical characteristics were analyzed, it was found their distributions did not conform normal distribution, with large variation coefficients. Clustering analysis then conducted TPI FPI within training group using , six categories classified. Subsequently, mapping relationship cutterhead speed, advance total force, established KNN classification model. It revealed when K-value set 4, has high macro - F 1 P R . Validated by testing data, method been proven be feasible effective. results indicate can effectively tunneling surrounding shield construction, particularly at face uniform homogeneous. This provides theoretical basis technical support for safe efficient tunneling.

Язык: Английский

Prediction of maximum dynamic shear modulus of undisturbed marine soils in the eastern coast of China based on machine learning methods DOI
Yiliang Tu, Qianglong Yao,

Ying Zhou

и другие.

Ocean Engineering, Год журнала: 2025, Номер 321, С. 120382 - 120382

Опубликована: Янв. 20, 2025

Язык: Английский

Процитировано

1
Recognizing Metaphorical Expressions in Chinese Speech and Their Natural Language Processing Strategies DOI Open Access
Zhuo Wang

Applied Mathematics and Nonlinear Sciences, Год журнала: 2025, Номер 10(1)

Опубликована: Янв. 1, 2025

Abstract Metaphor, as a special phenomenon in natural language, is of great significance for language processing tasks such sentiment analysis, machine translation, and question answer systems. In this paper, we design model metaphor recognition based on grammatical structure word meaning Chinese speech. The combines several key techniques recognition, using the TP-IDF algorithm feature extraction speech text, Bi-LSTM central network model. Finally, performance paper’s recognizing metaphors analyzed through an experimental design. When number layers attention mechanism 4, Precision, Recall, F1 are 94.32%, 95.03%, 93.36% respectively, effect optimal. TF-IDF adapts well to constructed paper. paper has good five types emotions except “surprise”, F-value MI_SS+MI_WS improved by 12.92%~26.18% compared with comparison method. This study promotes development provides new perspectives strategies other processing.

Язык: Английский

Процитировано

0
Statistical Analysis and Probabilistic Modeling of Shield Tunneling Data of Urban Large-Diameter Tunnel DOI Open Access
Hongyong Zhao, Limei Ran, Jie Li

и другие.

Journal of Physics Conference Series, Год журнала: 2025, Номер 3005(1), С. 012011 - 012011

Опубликована: Май 1, 2025

Abstract In the construction of underground spaces, shield tunneling method is increasingly adopted due to its advantages minimal surface disturbance and immunity external weather conditions. The setting parameters during crucial, as unreasonable parameter settings can reduce tunnel quality even lead risks such collapse construction. Most current studies focus on relationship between factors ground settlement geological types. This study takes a in Hangzhou case example, where were selected analyzed through mathematical statistics correlation analysis. normal variation range for different soil layers was determined. results showed that has certain with distribution strata. Total thrust negatively correlated speed, while it positively cutterhead torque. reason when strength front high, total torque increase, rotational speed decrease.

Язык: Английский

Процитировано

0
An intelligent method for leakage segmentation and area quantification to evaluate the safety performance of tunnels DOI
Tao Tian, Dechun Lu, Fanchao Kong

и другие.

Engineering Structures, Год журнала: 2025, Номер 338, С. 120581 - 120581

Опубликована: Май 20, 2025

Язык: Английский

Процитировано

0
Classification and prediction of rock mass drillability for a tunnel boring machine based on operational data mining DOI Creative Commons

Mingshe Sun,

Song Chen,

Huafei He

и другие.

Frontiers in Earth Science, Год журнала: 2024, Номер 12

Опубликована: Дек. 23, 2024

Currently, the accurate prediction of tunnel boring machine (TBM) performance remains a considerable challenge due to complex interactions between TBM and rock mass. In this study, research work is based on part metro project that covers 2,083.94 m. The Gaussian mixture model (GMM) K-nearest neighbor algorithm (KNN) are used classify predict mass drillability in excavation process. Drillability indexes introduced cluster mass, including penetration (P), field index (FPI), torque (TPI), specific energy (SE). Statistical characteristics were analyzed, it was found their distributions did not conform normal distribution, with large variation coefficients. Clustering analysis then conducted TPI FPI within training group using , six categories classified. Subsequently, mapping relationship cutterhead speed, advance total force, established KNN classification model. It revealed when K-value set 4, has high macro - F 1 P R . Validated by testing data, method been proven be feasible effective. results indicate can effectively tunneling surrounding shield construction, particularly at face uniform homogeneous. This provides theoretical basis technical support for safe efficient tunneling.

Язык: Английский

Процитировано

0