Time series prediction of shield construction ground settlement based on the fusion method of VMD-CNN-BiLSTM model and attention mechanism
GUO Junhao1, YAN Mengyao1, LI Jie1, ZHOU Lei1, ZHAO Xing1, HUANG Fu2, REN Ziheng2
1. China Railway 16th Bureau Group Corporation Limited, Beijing 100018, China; 2. School of Civil and Environmental Engineering, Changsha University of Science & Technology, Changsha 410114, Hunan, China
Abstract: In surface settlement prediction induced by shield tunneling, traditional empirical formulas and conventional single machine learning models often fell short in adequately capturing the nonlinear spatiotemporal characteristics embedded in multi-source parameters. To address this challenge, a deep learning-based predictive model was proposed in this research for effectively representing the dynamic interdependencies of stratum disturbance responses under complex geological conditions. Specifically, a hybrid VMD-CNN-BiLSTM-Attention framework was constructed by synergistically integrating variational mode decomposition(VMD), convolutional neural network(CNN), bidirectional long short-term memory(BiLSTM), and the attention mechanism. Within the proposed framework, VMD was first applied to decompose the raw settlement time series into multiple intrinsic mode functions, thereby facilitating noise reduction and latent pattern extraction. Subsequently, CNN was employed to extract spatial correlation features from both shield operational parameters(e.g., thrust, torque)and geotechnical properties(e.g., elastic modulus, cohesion, permeability coefficient). The high-level feature representations were then processed by the BiLSTM module to model bidirectional long-term temporal dependencies, while the attention mechanism was utilized to dynamically assign adaptive weights to salient time steps, thus enhancing both predictive accuracy and model interpretability. The model was trained and validated using field-monitored surface settlement data obtained during the shield-driven construction of the eastern extension of Changsha Metro Line 6, where the tunnel was successfully underpassed a critical airport runway. Comparative evaluation against in-situ measurements demonstrated that the proposed model achieved a low root mean square error(ERMS)on the test set and a high coefficient of determination(R2>0.96), which significantly outperformed conventional approaches in prediction accuracy. The results confirmed that the VMD-CNN-BiLSTM-Attention framework provided a robust and reliable tool for real-time settlement forecasting, thereby offering strong theoretical support for settlement control and early risk warning in shield tunneling projects.
郭军浩,颜梦瑶,李杰,周雷,赵兴,黄阜,任子衡. 基于VMD-CNN-BiLSTM模型与注意力机制融合的盾构施工诱发地表沉降时序预测[J]. 隧道与地下工程灾害防治, 2026, 8(1): 88-98.
GUO Junhao, YAN Mengyao, LI Jie, ZHOU Lei, ZHAO Xing, HUANG Fu, REN Ziheng. Time series prediction of shield construction ground settlement based on the fusion method of VMD-CNN-BiLSTM model and attention mechanism. Hazard Control in Tunnelling and Underground Engineering, 2026, 8(1): 88-98.
[1] LIN K, SUN Y Z, WANG J C, et al. Dynamic risk forecasting based on deep learning and collapse risk comprehensive evaluation of mountain tunnel portal construction[J]. Arabian Journal for Science and Engineering, 2025, 50(11): 8547-8565. [2] PAN Y, WANG Z Y, SUN L, et al. Dynamic prediction and multi-objective optimization on driving position of tunnel boring machine(TBM): an automated deep learning approach[J]. Acta Geotechnica, 2024, 19(8): 5611-5636. [3] ZHANG H L, WU Y X, CHENG J L, et al. A deep learning-based surrogate model for probabilistic analysis of high-speed railway tunnel crown settlement in spatially variable soil considering construction process[J]. Engineering Applications of Artificial Intelligence, 2024, 134: 108752. [4] XU C, CAO B T, YUAN Y, et al. A multi-fidelity deep operator network(DeepONet)for fusing simulation and monitoring data: application to real-time settlement prediction during tunnel construction[J]. Engineering Applications of Artificial Intelligence, 2024, 133: 108156. [5] 宋克志, 王梦恕, 孙谋. 基于 Peck 公式的盾构隧道地表沉降的可靠性分析[J]. 北京交通大学学报, 2004, 28(4): 30-33. SONG Kezhi, WANG Mengshu, SUN Mou. Reliability analysis of ground settlement with shield tunnel construction based on peck formula[J]. Journal of Beijing Jiaotong University, 2004, 28(4): 30-33. [6] PECK B B. Deep excavation and tunnelling in soft ground, state of the art volume[C] //7th ICSMFE. 1969, 4: 225-290. [7] 应宏伟, 熊一帆, 李冰河, 等. 软黏土基坑开挖诱发坑外地表沉降的时效解[J]. 岩土工程学报, 2024, 46(10): 2041-2050. YING Hongwei, XIONG Yifan, LI Binghe, et al. Time-dependent solution for ground settlement induced by excavation in soft clay[J]. Chinese Journal of Geotechnical Engineering, 2024, 46(10): 2041-2050. [8] 孙钧, 刘洪洲. 交叠隧道盾构法施工土体变形的三维数值模拟[J]. 同济大学学报(自然科学版), 2002, 30(4): 379-385. SUN Jun, LIU Hongzhou. 3-D numerical simulation of ground surface settlement under overlapped shield tunneling[J]. Journal of Tongji University(Natural Science), 2002, 30(4): 379-385. [9] 王儒, 翟五洲, 倪海波, 等. 盾构隧道机械法联络通道破洞施工中管片衬砌洞门结构力学响应的数值模拟研究[J]. 隧道建设(中英文), 2023, 43(增刊1): 178-188. WANG Ru, ZHAI Wuzhou, NI Haibo, et al. Numerical simulation on mechanical response of tunnel portal with segment lining in mechanized construction of cross passage of shield tunnel[J]. Tunnel Construction, 2023, 43(Suppl.1): 178-188. [10] 潘秋景, 李晓宙, 黄杉, 等. 机器学习在盾构隧道智能施工中的应用——综述与展望[J]. 隧道与地下工程灾害防治, 2022, 4(3): 10-30. PAN Qiujing, LI Xiaozhou, HUANG Shan, et al. Application of machine learning to intelligent shield tunnelling: review and prospects[J]. Hazard Control in Tunnelling and Underground Engineering, 2022, 4(3): 10-30. [11] FREITAG S, CAO B T, NINIC J, et al. Recurrent neural networks and proper orthogonal decomposition with interval data for real-time predictions of mechanised tunnelling processes[J]. Computers & Structures, 2018, 207: 258-273. [12] 熊文, 李宏洋, 傅鹤林, 等. 基于AJSO-BP的盾构施工地表变形预测[J]. 地下空间与工程学报, 2024, 20(增刊2): 949-955. XIONG Wen, LI Hongyang, FU Helin, et al. Prediction of surface deformation caused by shield tunneling based on AJSO-BP[J]. Chinese Journal of Underground Space and Engineering, 2024, 20(Suppl.2): 949-955. [13] SUWANSAWAT S, EINSTEIN H H. Artificial neural networks for predicting the maximum surface settlement caused by EPB shield tunneling[J]. Tunnelling and Underground Space Technology, 2006, 21(2): 133-150. [14] 傅蕾, 吴惠明, 黄宏伟, 等. 基于随机森林和MOPSO-CD的盾构隧道掘进沉降预测与施工参数优化[J]. 隧道建设(中英文), 2025, 45(增刊1): 318-329. FU Lei, WU Huiming, HUANG Hongwei, et al. Settlement prediction and construction parameters optimization induced by shield tunneling based on random forest and multi-objective particle swarm optimization algorithm-crowding distance[J]. Tunnel Construction, 2025, 45(Suppl.1): 318-329. [15] KONG F C, ZHOU X, GUO C X, et al. Elastic analytical method with machine learning for predicting the stratum displacement field induced by shallow tunneling[J]. Engineering Analysis with Boundary Elements, 2024, 159: 201-212. [16] 高修强, 彭达, 王国光, 等. 考虑盾构机参数主动控制的隧道掘进地表沉降智能预测方法[J]. 北京交通大学学报, 2024, 48(3): 120-129. GAO Xiuqiang, PENG Da, WANG Guoguang, et al. Intelligent prediction method for surface settlement in tunnel excavation considering active control of shield machine parameters[J]. Journal of Beijing Jiaotong University, 2024, 48(3): 120-129. [17] 王红卫, 周健, 简琦薇. 最近邻抽样回归模型在纵向地表沉降预测中的应用[J]. 土木工程学报, 2011, 44(增刊2): 226-229. WANG Hongwei, ZHOU Jian, JIAN Qiwei. Application of nearest neighbor bootstrap regressive model to forecasting the longitudinal ground settlement[J]. China Civil Engineering Journal, 2011, 44(Suppl.2): 226-229. [18] 殷宪辉, 牟朝群, 陈艺震, 等. 机理增强的盾构掘进过程横纵双向地表沉降联合智能预测[J]. 工业工程与管理, 2025, 30(2): 126-140. YIN Xianhui, MU Zhaoqun, CHEN Yizhen, et al. Mechanism-enhanced intelligent joint prediction of horizontal and longitudinal surface settlement in tunneling process[J]. Industrial Engineering and Management, 2025, 30(2): 126-140. [19] ZHANG P, WU H N, CHEN R P, et al. A critical evaluation of machine learning and deep learning in shield-ground interaction prediction[J]. Tunnelling and Underground Space Technology, 2020, 106: 103593. [20] 李洛宾, 龚晓南, 甘晓露, 等. 基于循环神经网络的盾构隧道引发地面最大沉降预测[J]. 土木工程学报, 2020, 53(增刊1): 13-19. LI Luobin, GONG Xiaonan, GAN Xiaolu, et al. Prediction of maximum ground settlement induced by shield tunneling based on recurrent neural network[J]. China Civil Engineering Journal, 2020, 53(Suppl.1): 13-19. [21] ZHANG N, ZHOU A N, PAN Y T, et al. Measurement and prediction of tunnelling-induced ground settlement in karst region by using expanding deep learning method[J]. Measurement, 2021, 183: 109700. [22] LECUN Y, BOTTOU L, BENGIO Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324. [23] GRAVES A, SCHMIDHUBER J. Framewise phoneme classification with bidirectional LSTM and other neural network architectures[J]. Neural Networks, 2005, 18(5/6): 602-610. [24] BAHDANAU D, CHO K, BENGIO Y. Neural machine translation by jointly learning to align and translate [J/OL]. Computer Science, 1-15[2014-09-01]. https://arxiv.org/abs/1409.0473 [25] CHICCO D, WARRENS M J, JURMAN G. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation[J]. PeerJ Computer Science, 2021, 7: e623. [26] IOFFE S, SZEGEDY C. Batch normalization: accelerating deep network training by reducing internal covariate shift[J/OL]. JMLR. ORG, 1-11[2025-02-11]. https://arxiv.org/abs/1502.03167 [27] 孟飞, 郑卓然, 黄文聪, 等. 基于CNN-BiLSTM-Attention的深基坑变形预测方法[J]. 地下空间与工程学报, 2025, 21(增刊1): 87-94. MENG Fei, ZHENG Zhuoran, HUANG Wencong, et al. Prediction method of pit deformation based on CNN-BiLSTM-attention[J]. Chinese Journal of Underground Space and Engineering, 2025, 21(Suppl.1): 87-94.
2026, Vol. 8 
