盾构下穿既有隧道影响控制及应用研究

于警举

摘 要

随着城市的发展，越来越多的城市开始向地下“要空间”。地下轨道交通在我国呈现出迅猛发展的态势，而地下空间通常是有限的，因此越来越多的工程面临着下穿既有隧道的问题。该类工程通常风险极大，新隧道的开挖会使既有隧道产生相应的沉降及水平位移，如不能对其扰动影响加以控制，轻则影响既有隧道的正常运营，重则破坏既有隧道的结构，带来较大的安全隐患。

基于此背景，本文结合安全预警项目实际下穿工程的需求，主要对通过优化施工参数来控制下穿过程的变形及沉降进行研究，从而解决实际的工程问题、为类似的工程提供指导和借鉴。主要研究内容及成果有以下几点：

（1）系统总结了盾构下穿既有隧道对地表及既有隧道所造成的扰动影响，并对新建隧道—土体—既有隧道三者之间的相互耦合影响机理进行总结和分析，对下穿的影响因素进行了归纳。

（2）以实际的下穿工程为例，阐述了利用数值模拟法、BP神经网络法优化施工参数的基本原理。分析了管棚超前支护方案无法实施的情况下，通过调整施工参数完成下穿过程的重要性及意义。

（3）详细阐述了利用试验段初步确定下穿施工参数的方法，基于此，利用Midas数值模拟软件对施工参数进行进一步的优化，对土仓压力、注浆压力、注浆量三个施工参数做出优化调整。此外，利用数值模拟软件对地表沉降及既有隧道沉降变形的规律进行了分析总结，并对该类下穿工程提出了相应的施工控制措施，为类似的工程提供参考借鉴。

（4）针对数值模拟法无法优化的施工参数，引入BP神经网络法，对下穿期间的施工参数进行优化。通过对既有隧道沉降的预测，实现施工参数的动态调整，从而指导实际的工程施工。通过试验段—数值模拟—BP神经网络三种方法对施工参数进行优化，最终下穿施工顺利完成。

关键词： 盾构下穿施工；既有隧道；施工参数优化；Midas；BP神经网络

Abstract

With the development of cities, more and more cities begin to develop underground space.Underground rail transit is developing rapidly in our country, but the underground space is usually limited, so more and more projects are faced with the problem of passing through existing tunnels. Such projects are usually very risky, and the excavation of new tunnels will cause settlement and horizontal displacement of existing tunnels,if the disturbance influence cannot be controlled, the normal operation of the existing tunnel will be affected, and even the structure of the existing tunnel will be damaged, which will bring greater safety hazards.

Based on this background, combined the actual engineering of the safety early warning project, this paper mainly studies the deformation and settlement controlled by optimizing construction parameters, so as to solve practical engineering problems and provide guidance and reference for similar projects. The main research contents and results are as follows:

（1）This paper systematically summarizes the disturbance effects of shield tunnels running through existing tunnels on the ground surface and existing tunnels, summarizes and analyzes the interaction mechanism of new tunnel-soil mass-existing tunnel, and summarized the influencing factors.

（2）Taking the actual underpass engineering as an example, this paper expounds the basic principle of optimizing the construction parameters by using the numerical simulation method and the BP neural network method. And analyzes the importance and significance of completing the underpass process by adjusting the construction parameters when the pipe shed advanced support scheme cannot be implemented.

（3）In this paper, the method of using the test section to initially determine the construction parameters of the underpass is described in detail. Based on this, the Midas numerical simulation software was used to further optimize the construction parameters, optimizing the three construction parameters of soil bin pressure, grouting pressure and grouting amount. In addition, the law of surface settlement and existing tunnel settlement was analyzed by numerical simulation software, and construction control measures for this type of underpass project are proposed, providing reference for similar projects.

（4）For the construction parameters that cannot be optimized by the numerical simulation method, this paper introduces the BP neural network method to optimize the construction parameters. By predicting the settlement of existing tunnels, the dynamic adjustment of construction parameters can be realized, so as to guide the actual engineering construction. The construction parameters were optimized by using the three methods of test section - numerical simulation - BP neural network, and finally the underpass construction was successfully completed.

Key words: Shield tunnel underpass construction; Existing tunnel；Optimization of construction parameters; Midas; BP neural network