基于深度学习的高寒区隧道衬砌应变预测方法

车博文; 包卫星; 卢汉青; 潘振华; 尹严

doi:10.20174/j.JUSE.2025.05.33

地下空间与工程学报 >

2025 , Vol. 21 >Issue 5: 1771 - 1783

DOI: https://doi.org/10.20174/j.JUSE.2025.05.33

防灾与环境

基于深度学习的高寒区隧道衬砌应变预测方法

车博文 ,
包卫星 ,
卢汉青 ,
潘振华 ,
尹严

展开

长安大学公路学院,西安 710064

车博文(1999—),男,宁夏银川人,硕士生,主要从事寒区隧道工程领域的科研工作。E-mail:cbwcbwcbw2@126.com

包卫星(1979—),男,乌鲁木齐人,博士,教授、博士生导师,主要从事特殊地区公路工程领域的科研工作。E-mail:baowx@chd.edu.cn

收稿日期: 2024-11-06

网络出版日期: 2025-10-17

基金资助

新疆重大科技专项(2020A03003-7);陕西省自然科学基础研究计划面上项目(2021JM-180);中央高校基本科研业务费资助项目(领军人才计划)(300102211302)

收起

Prediction Method of Tunnel Lining Strain in High Cold Regions Based on Deep Learning

Che Bowen ,
Bao Weixing ,
Lu Hanqing ,
Pan Zhenhua ,
Yin Yan

Expand

School of Highway, Chang' an University, Xi'an 710064, P.R. China

Received date: 2024-11-06

Online published: 2025-10-17

Fold

摘要

高寒区隧道衬砌结构所处的服役环境通常都较恶劣,容易受各种极端因素的耦合影响,导致衬砌混凝土产生不同形式的裂缝,给隧道结构带来了较大的安全隐患,因此对于高寒区隧道衬砌结构开裂行为的监测和预测具有重要工程意义,故提出了基于深度学习的高寒区隧道衬砌应变多因素耦合预测模型。以新疆某高寒区隧道为依托工程,以LSTM模型为基础,构建了一种CNN-Bilstm-Transformer (CBLT) 多因素耦合预测模型;基于大量现场监测数据对CBLT进行了训练与测试,探究了不同输入特征对模型预测性能的影响,并对比分析了CBLT与不同模型之间的预测性能。结果表明:随着输入特征的完善,CBLT的预测性能也逐渐提高,在测试集上的均方根误差(δ_RMSE)、平均绝对误差(δ_MAE)、平均绝对百分比误差(δ_MAPE)仅有2.61 με、2.22 με和1.24%,R²可达0.815,且预测性能优于RNN、LSTM、CNN-Bilstm和Bilstm-Transformer;将CBLT应用于实际工程中的衬砌开裂行为预测,取得了较高的精度。研究成果可为高寒区隧道衬砌开裂行为分析和冻害防治提供理论指导。

关键词： 高寒区隧道; 结构健康监测; 衬砌应变预测; 深度学习; 长短期记忆神经网络

本文引用格式

车博文 , 包卫星 , 卢汉青 , 潘振华 , 尹严 . 基于深度学习的高寒区隧道衬砌应变预测方法[J]. 地下空间与工程学报, 2025 , 21(5) : 1771 -1783 . DOI: 10.20174/j.JUSE.2025.05.33

Abstract

The service environment in which the lining structure of tunnels in high cold regions was usually harsh and susceptible to the coupling effects of various extreme factors, leading to different forms of cracks in the lining concrete and posing significant safety hazards to the entire tunnel structure. Therefore, monitoring and predicting the cracking behavior of tunnel lining structures in high cold regions has important engineering significance, so, a multi-factor coupling prediction model for tunnel lining strain in cold regions based on deep learning is proposed. Firstly, a CNN-Bilstm-Transformer (CBLT) multi-factor coupling prediction model was developed based on the LSTM model, using a high cold region tunnel of Xinjiang as a supporting project. Then, based on a large amount of on-site monitoring data, CBLT was trained and tested to explore the influence of different input features on the model's predictive performance, and the predictive performance between CBLT and different models was compared and analyzed. The research results indicate that: With the improvement of input features, the predictive performance of CBLT gradually improves; the average RMSE, MAE and MAPE on the test set were only 2.61 με, 2.22 με and 1.24%, R² can reach 0.815; the predictive performance of CBLT is better than RNN, LSTM, CNN-Bilstm and Bilstm-Transformer. The application of CBLT in predicting lining cracking behavior in practical engineering could achieve high accuracy. The research results can provide scientific guidance for the analysis of cracking behavior of tunnel lining in high cold regions and the prevention of frost damage.

Key words： high cold region tunnels; structural health monitoring; lining strain prediction; deep learning; long short-term memory neural network

参考文献

[1] 傅金阳,赵宁宁,肖欧辉,等.寒区隧道洞口仰拱混凝土早期开裂机理研究[J].地下空间与工程学报,2021, 17(4):1298-1308.(Fu Jinyang, Zhao Ningning, Xiao Ouhui, et al. Cracking mechanism of fill concrete at entrance section in cold region [J]. Chinese Journal of Underground Space and Engineering, 2021, 17(4): 1298-1308. (in Chinese))
[2] Taheri F, Shadlou S, Esmaeel R. Computational modelling of delamination and disbond in adhesively bonded joints and the relevant damage detection approaches [J]. Reviews of Adhesion anAdhesives, 2013, 1(4): 413-458.
[3] 许鹏, 伍毅敏, 张涛, 等.严寒地区高铁隧道衬砌挂冰机制及预测预警[J].铁道学报, 2023, 45(6): 169-174.(Xu Peng, Wu Yimin, Zhang Tao, et al.Ice hanging mechanism and early warning in high-speed railway tunnels in severe cold area[J].Journal of the China Railway Society, 2023,45(6):169-174. (in Chinese))
[4] 张素磊,张顶立,刘昌.公路运营隧道衬砌裂缝长期监测及分析[J].现代隧道技术,2017,54(3):17-25. (Zhang Sulei, Zhang Dingli, Liu Chang. Research on the design parameters of a single-track railway tunnel lining under local water pressure [J]. Modern Tunnelling Technology, 2017,54(3):17-25. (in Chinese))
[5] 李明, 陈卫忠, 杨建平, 等.基于功效系数法的隧道结构健康监测系统预警研究[J].岩土力学, 2015, 36(增2): 729-736.(Li Ming, Chen Weizhong, Yang Jianping, et al. Early warning research for tunnel structure health monitoring system based on efficacy coefficient method[J].Rock and Soil Mechanics, 2015,36(Supp.2):729-736. (in Chinese))
[6] 严晓东, 伍毅敏, 许鹏, 等.寒区隧道衬砌挂冰病害成因及预测预警技术[J].中国安全科学学报, 2019, 29(增2): 174-180.(Yan Xiaodong, Wu Yimin, Xu Peng, et al.Study on causes of frost damage of ice covering on lining in cold region tunnelsand forecast and early warning technology[J].China Safety Science Journal, 2019, 29 (Supp.2):174-180. (in Chinese))
[7] 徐进,林良宇,章龙管,等.基于深度学习的盾构掘进姿态预测模型[J].地下空间与工程学报,2022,18(增2):813-821.(Xu Jin, Lin Liangyu, Zhang Longguan, et al. Prediction model of shield tunneling attitude based on deep learning[J]. Chinese Journal of Underground Space and Engineering, 2022,18(Supp.2):813-821. (in Chinese))
[8] Du B W, Li W T, Tan X Y, et al. Development of load-temporal model to predict the further mechanical behaviors of tunnel structure under various boundary conditions [J]. Tunnelling and Underground Space Technology, 2021, 116: 104077.
[9] 马佳骥,马春驰,曾俊,等.基于LSTM多变种模型的岩爆微震参数预测研究[J].地下空间与工程学报,2022,18(5):1481-1494.(Ma Jiaji, Ma Chunchi, Zeng Jun, et al. Research on the prediction of rockburst microseismic parameters based on LSTM multivariate Model[J]. Chinese Journal of Underground Space and Engineering, 2022,18(5):1481-1494. (in Chinese))
[10] 何嵘国, 张孝勇, 汪旭, 等.基于深度学习的地铁隧道监控量测数据分析预测方法[J].隧道建设(中英文), 2021, 41(增2): 261-267.(He Rongguo, Zhang Xiaoyong, Wang Xu, et al. Analysis and prediction method for metro tunnel monitoring data based on deep learning[J].Tunnel Construction, 2021,41(Supp.2):261-267. (in Chinese))
[11] 陈城,史培新,贾鹏蛟,等.基于MK-LSTM算法的盾构掘进参数相关性分析及结构变形预测[J].吉林大学学报(工学版),2024,54(6):1624-1633.(Chen Cheng, Shi Peixin, Jia Pengjiao, et al. Correlation analysis of shield driving parameters and structural deformation prediction based on MK-LSTM algorithm[J].Journal of Jilin University (Engineering and Technology Edition), 2024, 54(6):1624-1633. (in Chinese))
[12] Tan X Y, Chen W Z, Tan X J, et al. Prediction for the future mechanical behavior of underwater shield tunnel fusing deep learning algorithm on SHM data [J]. Tunnelling and Underground Space Technology, 2022, 125: 104504.
[13] Xu P, Wu Y M, Wang Z J, et al. Distribution laws of freeze-thaw cycles and unsaturated concrete experiments in cold-region tunnels [J]. Cold Regions Science and Technology, 2020, 172: 102985.
[14] 陈建勋,荀文杰,赵鹏宇,等.考虑衬砌渐变劣化的寒区隧道承载性能演化规律[J].地下空间与工程学报,2023,19(5):1678-1690.(Chen Jianxun, Xun Wenjie, Zhao Pengyu, et al. Evolution law of bearing performance of tunnel in cold regions considering gradual deterioration of lining[J]. Chinese Journal of Underground Space and Engineering, 2023,19(5):1678-1690. (in Chinese))
[15] 潘振华,包卫星,郭强,等.高海拔寒区隧道水热力现场试验及数值模拟[J].地下空间与工程学报,2025,21(2):637-647.(Pan Zhenhua,Bao Weixing,Guo Qiang,et al.Field test and numerical simulation of hydraulic-thermal-mechanical tunnel in high altitude cold area[J].Chinese Journal of Underground Space and Engineering,2025,21(2):637-647.(in Chinese))
[16] 严健,陈四来,何川,等.平导压入通风对寒区隧道风温场的影响性研究[J].地下空间与工程学报,2021,17(4):1328-1336.(Yan Jian,Cheng Silai,He Chuan,et al.The Influence of parallel heading forced ventilation on the wind and temperature field of tunnel in cold region[J].Chinese Journal of Underground Space and Engineering,2021,17(4):1328-1336.(in Chinese))
[17] Jiang K J, Han Q, Du X L, et al. Structural dynamic response reconstruction and virtual sensing using a sequence to sequence modeling with attention mechanism [J]. Automation in Construction, 2021, 131: 103895.
[18] Greff K, Srivastava R K, Koutnik J, et al. LSTM: A search space odyssey [J]. IEEE Transactions on Neural Networks and Learning Systems, 2016, 28(10): 2222-2232.
[19] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need [J]. arXiv, 2017, 30:03762.
[20] 申彦斌, 张小丽, 夏勇, 等.Bi-LSTM神经网络用于轴承剩余使用寿命预测研究[J].振动工程学报, 2021, 34(2): 411-420.(Shen Yanbin, Zhang Xiaoli, Xia Yong, et al. Bi-LSTM neural network for remaining useful life prediction of bearings[J].Journal of Vibration Engineering, 2021,34(2):411-420. (in Chinese))
[21] 周中, 张俊杰, 丁昊晖, 等.基于GA-Bi-LSTM的盾构隧道下穿既有隧道沉降预测模型[J].岩石力学与工程学报, 2023, 42(1): 224-234.(Zhou Zhong, Zhang Junjie, Ding Haohui, et al. Settlement prediction model of shield tunnel under-crossing existing tunnel based on GA-Bi-LSTM[J]. Chinese Journal of Rock Mechanics and Engineering, 2023,42(1):224-234. (in Chinese))
[22] 高昆,于思淏,许维青,等.基于Attention-ResNet-LSTM混合神经网络的盾构掘进速度预测新方法[J].隧道建设(中英文), 2023, 43(4): 592-601.(Gao Kun, Yu Sihao, Xu Weiqing, et al. Novel prediction method for shield advancing speed based on Attention-ResNet-LSTM hybrid neural network[J]. Tunnel Construction, 2023, 43(4): 592-601. (in Chinese))
[23] 唐永圣, 姚彦东, 沈国根, 等.基于分布式应变传感的混凝土梁监测方法研究[J].铁道科学与工程学报, 2021, 18(4): 976-982.(Tang Yongsheng, Yao Yandong, Shen Guogen, et al.Study of monitoring method of concrete beam with distributed strain sensing technology[J].Journal of Railway Science and Engineering, 2021,18(4):976-982. (in Chinese))

Options

文章导航

模态框（Modal）标题

摘要

本文引用格式

Abstract

参考文献