科技创新与应用2025，Vol.15Issue(18)：57-61,5.DOI:10.19981/j.CN23-1581/G3.2025.18.012
基于BP神经网络的深部复合岩体变形模式预测
梁小刚 1刘沂星 1杨硕2
作者信息
- 1. 山西煤炭运销集团晋能煤矿工程有限公司,太原 030032
- 2. 徐州工程学院 土木工程学院,江苏 徐州 221018
- 折叠
摘要
Abstract
Predicting the deformation patterns of tunnel surrounding rock is of significant theoretical and practical importance for guiding tunnel construction,dynamic protection,and ensuring the long-term safe operation of tunnels.To predict and classify the deformation patterns of deep composite rock masses,this study utilizes data from composite rock mass model tests and DIC(Digital Image Correlation)deformation analysis.A BP(Back Propagation)neural network is employed to establish the complex nonlinear relationship between the global deformation of the surrounding rock and the deformation patterns,thereby achieving intelligent prediction of the deformation modes.The results show that the coefficient of determination(R2)between the predicted values and the actual values is 0.996,indicating a high prediction accuracy.The trained neural network demonstrates a strong capability in predicting the deformation patterns of the surrounding rock.The research results can provide an effective method for predicting deformation and fracture in practical engineering.关键词
神经网络/复合岩体/变形破裂模式/DIC变形分析数据/模型训练Key words
neural network/composite rock mass/deformation and fracture mode/DIC deformation and analysis data/model training分类
土木建筑引用本文复制引用
梁小刚,刘沂星,杨硕..基于BP神经网络的深部复合岩体变形模式预测[J].科技创新与应用,2025,15(18):57-61,5.
