{"title":"Block Collapse Prediction and Reinforcement Optimization in Tunnels Based on Discontinuous Deformation Analysis and Machine Learning Models","authors":"Hongyun Fan, Liping Li, Yuguang Fu, Hongliang Liu, Xiangyu Chang, Xin Gao","doi":"10.1002/nag.70102","DOIUrl":null,"url":null,"abstract":"<jats:label/>Block collapse is one of the most common geological hazards encountered during tunnel construction, characterized by its sudden occurrence and severe consequences. Currently, the prediction and prevention of tunnel block collapse rely primarily on theoretical analysis, numerical simulations, and physical experiments. However, these approaches often oversimplify real‐world conditions and lack efficiency. This study proposes a prediction and reinforcement optimization method for block collapse by integrating discontinuous deformation analysis (DDA) with machine learning method. First, DDA method was employed to simulate tunnel block collapse under structural plane inclination angles of 15°, 30°, and 45°. The corresponding simulation errors compared to model test results were 2.68%, 3.76%, and 1.01%, respectively, demonstrating the accuracy of the DDA approach in modeling block collapse. Next, a dataset of 142 block collapse scenarios under varying conditions was established, encompassing multiple parameters. Among them, the spacing, inclination angle, and internal friction angle of structural planes were identified as having the most significant influence on collapse behavior. Subsequently, five machine learning models were developed to predict collapse height, affected area, and perimeter deformation. All models achieved high coefficients of determination (<jats:italic>R</jats:italic><jats:sup>2</jats:sup>), with XGBoost exhibiting the best performance. Finally, a data‐driven method for optimizing reinforcement parameters was introduced by integrating the DDA and XGBoost models. Based on prediction results for Cases 128–142, the number of required rock bolts was successfully reduced from 255 to 98 using the proposed optimization strategy. This research provides a valuable reference for the design of reinforcement measures in practical tunnel construction in rock masses.","PeriodicalId":13786,"journal":{"name":"International Journal for Numerical and Analytical Methods in Geomechanics","volume":"204 1","pages":""},"PeriodicalIF":3.6000,"publicationDate":"2025-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal for Numerical and Analytical Methods in Geomechanics","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1002/nag.70102","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, GEOLOGICAL","Score":null,"Total":0}
引用次数: 0
Abstract
Block collapse is one of the most common geological hazards encountered during tunnel construction, characterized by its sudden occurrence and severe consequences. Currently, the prediction and prevention of tunnel block collapse rely primarily on theoretical analysis, numerical simulations, and physical experiments. However, these approaches often oversimplify real‐world conditions and lack efficiency. This study proposes a prediction and reinforcement optimization method for block collapse by integrating discontinuous deformation analysis (DDA) with machine learning method. First, DDA method was employed to simulate tunnel block collapse under structural plane inclination angles of 15°, 30°, and 45°. The corresponding simulation errors compared to model test results were 2.68%, 3.76%, and 1.01%, respectively, demonstrating the accuracy of the DDA approach in modeling block collapse. Next, a dataset of 142 block collapse scenarios under varying conditions was established, encompassing multiple parameters. Among them, the spacing, inclination angle, and internal friction angle of structural planes were identified as having the most significant influence on collapse behavior. Subsequently, five machine learning models were developed to predict collapse height, affected area, and perimeter deformation. All models achieved high coefficients of determination (R2), with XGBoost exhibiting the best performance. Finally, a data‐driven method for optimizing reinforcement parameters was introduced by integrating the DDA and XGBoost models. Based on prediction results for Cases 128–142, the number of required rock bolts was successfully reduced from 255 to 98 using the proposed optimization strategy. This research provides a valuable reference for the design of reinforcement measures in practical tunnel construction in rock masses.
期刊介绍:
The journal welcomes manuscripts that substantially contribute to the understanding of the complex mechanical behaviour of geomaterials (soils, rocks, concrete, ice, snow, and powders), through innovative experimental techniques, and/or through the development of novel numerical or hybrid experimental/numerical modelling concepts in geomechanics. Topics of interest include instabilities and localization, interface and surface phenomena, fracture and failure, multi-physics and other time-dependent phenomena, micromechanics and multi-scale methods, and inverse analysis and stochastic methods. Papers related to energy and environmental issues are particularly welcome. The illustration of the proposed methods and techniques to engineering problems is encouraged. However, manuscripts dealing with applications of existing methods, or proposing incremental improvements to existing methods – in particular marginal extensions of existing analytical solutions or numerical methods – will not be considered for review.