融合随机介质理论与深度学习的隧道复杂收敛模式确定:以广花城际盾构工程为例
DOI:
CSTR:
作者:
作者单位:

北京科技大学

作者简介:

通讯作者:

中图分类号:

TU91

基金项目:

国家重点研发计划课题(2023YFC3707803)


Determination of Complex Tunnel Convergence Patterns by Integrating Stochastic Medium Theory and Deep Learning: Guanghua Intercity Railway Project
Author:
Affiliation:

University of Science and Technology Beijing

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    隧道收敛模式的确定是隧道开挖引起地层变形分析的关键环节,其固有的隐蔽性和复杂性使其难以通过现场监测和理论分析得到,现有的收敛模式往往基于工程经验和类比分析得到,这限制了地层变形分析结果的精度。为了建立隧道复杂收敛模式的确定方法,首先依据三个独立坐标系的变换关系,结合傅里叶级数,建立了隧道收敛模式的统一方程,实现了对隧道收敛模式的准确描述;随后基于随机介质理论开展了参数敏感性分析,研究了不同收敛模式参数对地表沉降的影响规律;进一步基于随机介质理论,通过高维网格离散得到了样本集,并以地层变形为输入,收敛模式统一方程的参数为输出建立数据映射,从而实现了复杂收敛模式的反演确定;最后结合工程实例验证了方法的可靠性。研究结果表明:不均匀收敛模式虽被工程行业广泛接受,但其对隧道轮廓变形的假设过于简化,并不符合实际;复杂施工条件下,变形后的隧道轮廓并非严格位于开挖空间内部,地层损失区的分布也可能并不对称,这将使得地表沉降曲线的峰值并不严格位于开挖轮廓中心的正上方。实际工程中,隧道收敛模式的影响因素较为复杂,该方法可通过简单的现场实测变形数据实现隧道收敛模式的反演确定,能够为隧道开挖引起地层变形的精细化分析和双擎预测提供依据。

    Abstract:

    The determination of tunnel convergence patterns is a critical step in the analysis of strata deformation induced by tunnel excavation. However, due to their inherent concealment and complexity, tunnel convergence patterns are difficult to accurately determine through field monitoring and theoretical analysis. Existing convergence patterns are often derived from engineering experience and analogy-based methods, which limit the accuracy of strata deformation analysis. To develop a method for determining complex tunnel convergence patterns, a unified equation for the convergence patterns was established based on the transformation relationship of three independent coordinate systems and Fourier series, which enables an accurate description of such patterns. Subsequently, parameter sensitivity analysis was conducted based on stochastic medium theory to investigate the influence of different convergence pattern parameters on ground settlement. Then, a sample set was generated by discretizing the domain into a high-dimensional grid, and a data mapping was established using ground settlement as input and the parameters of the unified equation for the convergence patterns as output, thereby enabling the inversion of complex convergence patterns. Finally, the reliability of the method was validated via an engineering case study. The results indicate that although the non-uniform convergence pattern is widely adopted in engineering practice, its assumptions regarding tunnel contour deformation are oversimplified and may not accurately reflect actual conditions. Under complex geological and construction conditions, the deformed tunnel contour may not strictly remain within the excavation boundary, and the ground loss region may exhibit asymmetry, causing the peak of ground settlement to deviate from the tunnel centerline. In engineering practice, the factors influencing tunnel convergence patterns are highly complex. The proposed method can determine the tunnel convergence pattern using simple field-measured deformation data, thereby providing a basis for refined analysis and dual-engine prediction of strata deformation induced by tunnel excavation.

    参考文献
    相似文献
    引证文献
引用本文

赵佳乐,刘洋,王肖肖,等. 融合随机介质理论与深度学习的隧道复杂收敛模式确定:以广花城际盾构工程为例[J]. 科学技术与工程, , ():

复制
分享
相关视频

文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2026-05-26
  • 最后修改日期:2026-08-27
  • 录用日期:2026-09-13
  • 在线发布日期:
  • 出版日期:
×
关于抵制伪造参考文献学术不端行为的声明
关于规范使用生成式人工智能的声明
喜报|《科学技术与工程》3篇论文成功入选 “第二十八届中国科协年会影响力提名论文”
2026年会通知 | “技术经济学驱动智能经济生态构建与治理变革”——中国技术经济学会第三十三届学术年会(2026)会议通知暨征文启事(第一轮)
亟待确认版面费归属稿件,敬请作者关注