Estimation model for building settlement around subway foundation pit based on spatiotemporal characteristics and deep learning
YAO Ronghan1，ZHANG Wensong2，3，4，JIA Lei2，3，4，WANG Libing1
(1. School of Transportation and Vehicle Engineering，Shandong University of Technology，Zibo，Shandong 255049，China；
2. School of Urban Geology and Engineering，Hebei GEO University，Shijiazhuang，Hebei 050031，China；3. Hebei
Technology Innovation Center for Intelligent Development and Control of Underground Built Environment，
Shijiazhuang，Hebei 050031，China；4. Key Laboratory of Intelligent Detection and Equipment for
Underground Space of Beijing—Tianjin—Hebei Urban Agglomeration，Ministry of
Natural Resources，Shijiazhuang，Hebei 050031，China)
Abstract：The excavation of subway foundation pits inevitably leads to the settlement of the surrounding ground surface and buildings. Therefore，real-time and accurate monitoring of ground surface and building settlement is crucial. To thoroughly investigate the spatiotemporal characteristics of building settlement resulting from subway foundation pit excavations and enhance the precision of settlement estimations，a convolutional gated recurrent unit neural network(ConvGRU) model for the building settlement estimation was proposed based on deep learning. The spatiotemporal matrix of the building settlement was built using the building settlement data obtained from the subject and adjacent monitoring points. The convolutional neural network(CNN) was employed to capture the spatial patterns of the building settlement data，while the gated recurrent unit(GRU) neural network was utilized to extract the temporal patterns. This dual approach allows for a comprehensive analysis of the spatiotemporal characteristics of the building settlement data. The estimation performance of the ConvGRU model was compared with that of the history average model and those of four existing deep learning models using the data of the building settlement around a subway foundation pit in Shenzhen city of Guangdong Province of China. The results indicate that the estimation error of the ConvGRU model is reduced by 12.86% to 39.00% when compared to the five existing models. This research demonstrates that the ConvGRU model achieves higher accuracy and better generalization，providing high-quality building settlement estimation data for subway construction and serving as a warning mechanism for ground surface and building settlement associated with subway foundation pits.
姚荣涵1，张文松2，3，4，贾 磊2，3，4，王立冰1. 基于时空特性和深度学习的地铁基坑周边建筑物沉降估算模型[J]. 岩石力学与工程学报, 2025, 44(S1): 196-205.
YAO Ronghan1，ZHANG Wensong2，3，4，JIA Lei2，3，4，WANG Libing1. Estimation model for building settlement around subway foundation pit based on spatiotemporal characteristics and deep learning. , 2025, 44(S1): 196-205.
[1] 钟国强，王 浩，李 莉，等. 基于SFLA-GRNN模型的基坑地表最大沉降预测[J]. 岩土力学，2019，40(2)：792–798.(ZHONG Guoqiang，WANG Hao，LI Li，et al. Prediction of maximum settlement of foundation pit based on SFLA-GRNN model[J]. Rock and Soil Mechanics，2019，40(2)：792–798.(in Chinese))
[2] 张 斌，顾功开，李 超. 基于动平均改进灰色模型的高边坡沉降变形预测[J]. 人民长江，2010，41(20)：56–59.(ZHANG Bin，GU Gongkai，LI Chao. Forecast for settlement of high slope based on improved dynamic average grey model[J]. Yangtze River，2010，41(20)：56–59.(in Chinese))
[3] 伍毅敏，刘延安，王 恒，等. 北京市隧道下穿施工引起城市路面沉降的影响规律回归分析[J]. 交通运输工程学报，2022，22(2)：176–186.(WU Yimin，LIU Yan?an，WANG Heng，et al. Regression analysis of influence law of urban pavement settlement caused by underpass tunnel construction in Beijing[J]. Journal of Traffic and Transportation Engineering，2022，22(2)：176–186.(in Chinese))
[4] HUANG C F，LI Q，WU S C，et al. Application of the Richards model for settlement prediction based on a bidirectional difference-weighted least-squares method[J]. Arabian Journal for Science and Engineering，2018，43(10)：5 057–5 065.
[5] YANG C W，JIANG Z Z. A discrete-time model-based method for predicting settlement of geotechnical foundations in buildings[J]. Mobile Information Systems，2022：5631634.
[6] 景宏君，苏如荣，苏 霆. 高路堤沉降变形预测模型研究[J]. 岩土力学，2007，139(8)：1 762–1 766.(JING Hongjun，SU Rurong，SU Ting. Study of settlement deformation prediction model of high embankment[J]. Rock and Soil Mechanics，2007，139(8)：1 762– 1 766.(in Chinese))
[7] CUI Z D，REN S X. Prediction of long-term settlements of subway tunnel in the soft soil area[J]. Natural Hazards，2014，74(2)：1 007– 1 020.
[8] 张文胜，崔志伟. 铁路客运专线特大桥沉降预测模型[J]. 交通运输工程学报，2011，11(6)：31–36.(ZHANG Wensheng，CUI Zhiwei. Settlement prediction model of super large bridge for passenger dedicated railway[J]. Journal of Traffic and Transportation Engineering，2011，11(6)：31–36.(in Chinese))
[9] 黄亚东，张土乔，俞亭超，等. 公路软基沉降预测的支持向量机模型[J]. 岩土力学，2005，26(12)：1 987–1 990.(HUANG Yadong，ZHANG Tuqiao，YU Tingchao，et al. Support vector machine model of settlement prediction of road soft foundation[J]. Rock and Soil Mechanics，2005，26(12)：1 987–1 990.(in Chinese))
[10] QIAO S F，TAN J K，ZHANG Y G，et al. Settlement prediction of foundation pit excavation based on the GWO-ELM model considering different states of influence[J]. Advances in Civil Engineering，2021，Article number：8896210.
[11] DENG J H，ZENG T，YUAN S，et al. Interval prediction of building foundation settlement using kernel extreme learning machine[J]. Frontiers in Earth Science，2022，10：8896210.
[12] ZHANG C，LI J Z，HE Y. Application of optimized grey discrete Verhulst-BP neural network model in settlement prediction of
foundation pit[J]. Environmental Earth Sciences，2019，78(15)： 441.
[13] ZHANG Z Y，XU R Q，WU X，et al. ANN-based dynamic prediction of daily ground settlement of foundation pit considering time- dependent influence factors[J]. Applied Sciences-Basel，2022，12(13)：6 324.
[14] 文 明，张顶立，房 倩，等. 地铁车站施工过程中地表沉降的NARXNN时间序列预测模型[J]. 岩石力学与工程学报，2015，34(增1)：3 306–3 312.(WEN Ming，ZHANG Dingli，FANG Qian，et al. The NARXNN time series prediction model for ground subsidence caused by construction of metro station[J]. Chinese Journal of Rock Mechanics and Engineering，2015，34(Supp.1)：3 306– 3 312.(in Chinese))
[15] 高 燕，吴晓东，田嘉逸. 基于机器学习的地下溶洞空间形态对地面沉降影响的分析与预测[J]. 中山大学学报：自然科学版(中英文)，2023，62(2)：83–92.(GAO Yan，WU Xiaodong，TIAN Jiayi. Analysis and prediction of the ground subsidence due to the spatial form of underground karst caves based on machine learning[J]. Acta Scientiarum Naturalium Universitatis Sunyatseni，2023，62(2)： 83–92.(in Chinese))
[16] 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))
[17] 武 焱，张映雪. 基于门控深度循环信念网络的边坡沉降预测[J]. 交通科学与工程，2023，39(1)：26–34.(WU Yan，ZHANG Yingxue. Slope settlement prediction based on gated deep recurrent belief network[J]. Journal of Transport Science and Engineering，2023，39(1)：26–34.(in Chinese))
[18] 于永堂，郑建国，张继文，等. 基于卡尔曼滤波与指数平滑法融合模型的沉降预测新方法[J]. 岩土工程学报，2021，43(增1)：127–131.(YU Yongtang，ZHENG Jianguo，ZHANG Jiwen，et al. Prediction of settlement based on fusion model of Kalman filter and exponential smoothing algorithm[J]. Chinese Journal of Geotechnical Engineering，2021，43(Supp.1)：127–131.(in Chinese))
[19] 马学宁，陈玉燕，王 旭. 高速铁路车站岔区高填方路基沉降组合预测研究[J]. 铁道学报，2023，45(1)：105–113.(MA Xuening，CHEN Yuyan，WANG Xu. Study on settlement combination prediction of high fill subgrade in turnout area of high-speed railway station[J]. Journal of the China Railway Society，2023，45(1)：105–113.(in Chinese))
[20] 李 涛，杨腾宇，刘 波，等. 基于VMD-GRU的地铁隧道台阶法施工地表沉降预测[J]. 华中科技大学学报：自然科学版，2023，51(7)：48–54.(LI Tao，YANG Tengyu，LIU Bo，et al. Surface settlement prediction of subway tunnels constructed by step method based on VMD-GRU[J]. Journal of Huazhong University of Science and Technology：Natural Science，2023，51(7)：48–54.(in Chinese))
[21] MA X L，DAI Z，HE Z B，et al. Learning traffic as images：a deep convolutional neural network for large-scale transportation network speed prediction[J]. Sensors，2017，17(4)：818.
[22] ZHANG D，KABUKA M R. Combining weather condition data to predict traffic flow：a GRU-based deep learning approach[J]. IET Intelligent Transport Systems，2018，12(7)：578–585.
[23] YAO R H，ZHANG W S，LONG M. DLW-Net model for traffic flow prediction under adverse weather[J]. Transportmetrica B-Transport Dynamics，2022，10(1)：499–524.
[24] 张文松，姚荣涵. 基于时空特性和组合深度学习的交通流参数估计[J]. 交通运输系统工程与信息，2021，21(1)：82–89.(ZHANG Wensong，YAO Ronghan. Traffic flow parameters estimation based on spatio-temporal characteristics and hybrid deep learning[J]. Journal of Transportation Systems Engineering and Information Technology，2021，21(1)：82–89.(in Chinese))
2025, Vol. 44 
