基于长短时记忆的真空预压地基沉降预测

doi:10.16183/j.cnki.jsjtu.2023.340

上海交通大学学报 ›› 2025, Vol. 59 ›› Issue (4): 525-532.doi: 10.16183/j.cnki.jsjtu.2023.340

基于长短时记忆的真空预压地基沉降预测

梁煜婉¹^,², 肖朝昀¹^,²^,³(), 李明广⁴, 孟江山⁵, 周建烽¹^,², 黄山景³, 朱浩杰³

1.华侨大学土木工程学院,福建厦门 361021
2.福建省智慧基础设施与监测重点实验室, 福建厦门 361021
3.华土木(厦门)科技有限公司,福建厦门 361000
4.上海交通大学船舶海洋与建筑工程学院,上海 200240
5.中交广州航道局有限公司,广州 510290

收稿日期:2023-07-24 修回日期:2023-10-27 接受日期:2023-11-17 出版日期:2025-04-28 发布日期:2025-05-09
通讯作者: 肖朝昀 E-mail:zyxiao@hqu.edu.cn
作者简介:梁煜婉(1998—),硕士生,从事地下工程与地基基础研究.
基金资助:
国家自然科学基金(52278361);福建省高校产学合作(2023Y4007)

Ground Settlement Prediction by Vacuum Preloading Based on LSTM

LIANG Yuwan¹^,², XIAO Zhaoyun¹^,²^,³(), LI Mingguang⁴, MENG Jiangshan⁵, ZHOU Jianfeng¹^,², HUANG Shanjing³, ZHU Haojie³

1. College of Civil Engineering, Huaqiao University, Xiamen 361021, Fujian, China
2. Key Laboratory for Intelligent Infrastructure and Monitoring of Fujian Province, Xiamen 361021, Fujian, China
3. China Civil Engineering (Xiamen) Technology Co., Ltd., Xiamen 361000, Fujian, China
4. School of Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
5. CCCC Guangzhou Dredging Co., Ltd., Guangzhou 510290, China

Received:2023-07-24 Revised:2023-10-27 Accepted:2023-11-17 Online:2025-04-28 Published:2025-05-09
Contact: XIAO Zhaoyun E-mail:zyxiao@hqu.edu.cn

摘要/Abstract

摘要：

为探寻一种更加准确的真空预压地基处理沉降预测方法,以厦门新机场规划片区东园地块造地二期工程为例,构建基于长短时记忆(LSTM)神经网络的真空预压地基处理沉降预测模型.选取两个区域的实测沉降数据作为数据基础,对比传统沉降预测法(浅岗法、三点法和双曲线法)与LSTM神经网络预测结果.研究结果表明:当真空预压地基处理工况下出现真空膜破损引发沉降量回弹的现象时,相较于传统预测方法,LSTM的均方根误差e_RMSE和平均绝对值误差e_MAE均下降45%以上,且该方法的预测结果有明显的上升趋势,能够准确预测出沉降回弹情况.在预测误差方面,考虑真空度和沉降变化的LSTM模型比仅考虑沉降时序的LSTM模型的e_RMSE和e_MAE降低60%及以上.该研究可为真空预压地基沉降预测提供先进的数据驱动预测方法.

关键词: 深度学习, 长短期记忆网络, 真空预压, 沉降预测

Abstract:

In order to explore a more accurate method for predicting settlement in vacuum preloading foundation treatment, a vacuum preloading settlement prediction model based on long short-term memory (LSTM) neural network was developed, taking the second-phase land reclamation project in the East Park of Xiamen New Airport planning area as an example. Measured settlement data from two regions were selected as the dataset, and the results were compared with traditional settlement prediction methods including the Asaoka method, three-point method, and hyperbolic method. The results show that the prediction model based on the LSTM neural network considering only sedimentation time series outerperforms the traditional methods that rely only on sedimentation time series. When the vacuum film is damaged and settlement rebound occurs under vacuum precompression foundation treatment, the root mean squared error (e_RMSE) and the mean absolute error (e_MAE) of LSTM model decrease by more than 45% compared to the traditional methods. Additionly, this model accurately captures the settlement rebound trend, providing more reliable prediction. In terms of prediction error, the e_RMSE and e_MAE values of the LSTM model which considers vacuum degree and sedimentation are lower than those of the LSTM model which only considers sedimentation time series by over 60%. This paper offers an advanced data-driven prediction method for prediction in vacuum preloading foundation settlement.

Key words: deep learning, long short term memory (LSTM), vacuum preloading, settlement prediction

中图分类号:

梁煜婉, 肖朝昀, 李明广, 孟江山, 周建烽, 黄山景, 朱浩杰. 基于长短时记忆的真空预压地基沉降预测[J]. 上海交通大学学报, 2025, 59(4): 525-532.

LIANG Yuwan, XIAO Zhaoyun, LI Mingguang, MENG Jiangshan, ZHOU Jianfeng, HUANG Shanjing, ZHU Haojie. Ground Settlement Prediction by Vacuum Preloading Based on LSTM[J]. Journal of Shanghai Jiao Tong University, 2025, 59(4): 525-532.

图/表 10

图1

图2

图3

图4

图5

表1

表2

图6

图7

表3

参考文献 13

[1]	SRIDHARAN A, MURTHY N S, PRAKASH K. Rectangular hyperbola method of consolidation analysis[J]. Géotechnique, 1987, 37(3): 355-368.
[2]	曾国熙, 杨锡令. 砂井地基沉陷分析[J]. 浙江大学学报, 1959(3): 34-72.
	ZENG Guoxi, YANG Xiling. Analysis of sand well foundation subsidence[J]. Journal of Zhejiang University, 1959(3): 34-72.
[3]	ASAOKA A. Observational procedure of settlement prediction[J]. Soils and Foundations, 1978, 18(4): 87-101.
[4]	潘林有, 谢新宇. 用曲线拟合的方法预测软土地基沉降[J]. 岩土力学, 2004, 25(7): 1053-1058.
	PAN Linyou, XIE Xinyu. Observational settlement prediction by curves fitting methods[J]. Rock and Soil Mechanics, 2004, 25(7): 1053-1058.
[5]	黄广军. Asaoka法预测软土地基沉降时存在的问题和对策[J]. 岩土力学, 2016, 37(4): 1061-1065.
	HUANG Guangjun. Problems and their solutions in predicting soft ground settlement based on Asaoka’s method[J]. Rock Soil Mechanics, 2016, 37(4): 1061-1065.
[6]	HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780. doi: 10.1162/neco.1997.9.8.1735 pmid: 9377276
[7]	XU Y, LU X, CETINER B, et al. Real-time regional seismic damage assessment framework based on long short-term memory neural network[J]. Computer-Aided Civil and Infrastructure Engineering, 2021, 36(4): 504-521.
[8]	SENANAYAKE S, PRADHAN B, ALAMRI A, et al. A new application of deep neural network (LSTM) and RUSLE models in soil erosion prediction[J]. Science of The Total Environment, 2022, 845: 157-220.
[9]	刘俊城, 谭勇, 张生杰. 地铁车站深基坑开挖变形智能多步预测方法[J]. 上海交通大学学报, 2024, 58(7): 1108-1117. doi: 10.16183/j.cnki.jsjtu.2022.419
	LIU Juncheng, TAN Yong, ZHANG Shengjie. Intelligent multi-step prediction method for deep foundation pit deformation of subway station[J]. Journal of Shanghai Jiao Tong University, 2024, 58(7): 1108-1117.
[10]	GAO B, WANG R R, LIN C, et al. TBM penetration rate prediction based on the long short-term memory neural network[J]. Underground Space, 2021, 6(6): 718-731.
[11]	HU Y, GU C, MENG Z, et al. Prediction for the settlement of concrete face rockfill dams using optimized LSTM model via correlated monitoring data[J]. Water, 2022, 14(14): 21-57.
[12]	王启贵, 吴忠明, 陈兴红. 基于机器学习LSTM方法的大面积高填土沉降预测[J]. 工程勘察, 2022, 50(8): 41-45.
	WANG Qigui, WU Zhongming, CHEN Xinghong. Large area high fill settlement prediction based on machine learning LSTM method[J]. Engineering Investigation and Investigation, 2022, 50(8): 41-45.
[13]	龚晓南, 岑仰润. 真空预压加固软土地基机理探讨[J]. 哈尔滨建筑大学学报, 2002(2): 7-10.
	GONG Xiaonan, CEN Yangrun. Discussion on mechanism of vacuum preloading strengthening soft soil foundation[J]. Journal of Harbin Jianzhu University, 2002(2): 7-10.

阶段	LSTM		三点法		双曲线法		浅岗法
阶段	e_RMSE/mm	e_MAE/mm	e_RMSE/mm	e_MAE/mm	e_RMSE/mm	e_MAE/mm	e_RMSE/mm	e_MAE/mm
拟合阶段	3.5	3.0	3.3	2.2	10.7	9.0	1.3	0.4
预测阶段	0.7	0.6	1.0	0.9	18.1	14.7	1.5	1.3

阶段	LSTM		三点法		双曲线法		浅岗法
阶段	e_RMSE/mm	e_MAE/mm	e_RMSE/mm	e_MAE/mm	e_RMSE/mm	e_MAE/mm	e_RMSE/mm	e_MAE/mm
拟合阶段	3.9	3.4	2.4	1.5	2.5	2.1	3.5	2.9
预测阶段	2.7	2.3	12.4	10.7	13.6	11.5	5.0	4.3

区域	只考虑沉降时序		考虑真空度和沉降时序
区域	e_RMSE/ mm	e_MAE/ mm	e_RMSE/ mm	e_MAE/ mm
B6-1-17	0.7	0.6	0.4	0.4
B4-1-6	2.7	2.3	0.6	0.9
B4-1-6(沉降回弹阶段)	3.4	3.1	1.1	0.8

[1]	. 基于改进YOLOv5l的交通信号灯识别[J]. J Shanghai Jiaotong Univ Sci, 2026, 31(2): 319-333.
[2]	郭琦, 闫军, 郝乾鹏, 韩东, 杨志豪, 闫馨月, 张海鹏, 李然. 基于闭环聚类和多目标优化的风电短期功率预测方法[J]. 上海交通大学学报, 2026, 60(2): 246-255.
[3]	陈亮汶, 朱宇昕, 沈涛, 俞羿帆, 凌霄, 盛庆红. 基于红外尾迹匹配的舰船目标检测算法[J]. 空天防御, 2026, 9(1): 80-90.
[4]	罗志军, 王健瑞, 殷佳伟. 复杂战场环境下的任务驱动智能目标识别方法综述[J]. 空天防御, 2026, 9(1): 1-11.
[5]	. 基于深度学习序列方法的多人姿态估计用来检测人体与关键点位置[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(6): 1103-1113.
[6]	荣光, 张业鑫, 唐朝, 陈金宝, 周奕玲, 王建园. 基于仿真数据驱动的无人飞行器故障诊断技术研究[J]. 空天防御, 2025, 8(6): 73-84.
[7]	. 基于三维卷积特征金字塔网络的高光谱卫星图像分类[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(6): 1073-1084.
[8]	. 基于CEEMDAN 和 GRU的停车位预测[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(5): 962-975.
[9]	. 基于ALBERT的中国诗酒文化命名实体识别[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(5): 1065-1072.
[10]	. 面向太阳能电池复杂缺陷检测的新型多步深度学习方法[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(5): 1050-1064.
[11]	夏伊琳, 刘刚, 鄢丛强, 蔡云泽. 基于深度学习的SAR图像舰船尾迹旋转框检测算法研究[J]. 空天防御, 2025, 8(5): 64-74.
[12]	许强, 马跃华, 许可, 潘俊. 雷达目标智能识别方法研究综述[J]. 空天防御, 2025, 8(5): 1-9.
[13]	赵紫昱, 王绪泉, 马杰, 邢裕杰, 顿雄, 王占山, 程鑫彬. 轻薄红外计算成像重建算法的边缘芯片部署方法研究[J]. 空天防御, 2025, 8(4): 85-93.
[14]	. 基于表面肌电信号的BP-LSTM混合模型肘部运动实时预测[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(3): 455-462.
[15]	. 基于RGB-D图像的机器人抓取检测高效全卷积网络和优化方法[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(2): 399-416.

基于长短时记忆的真空预压地基沉降预测

Ground Settlement Prediction by Vacuum Preloading Based on LSTM

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 13

相关文章 15

编辑推荐

Metrics

本文评价