摘 要

   近年来，随着我国城市化进程的加快，给城市交通及土地资源带来了巨大的挑战，地铁建设成为解决有效利用城市空间、满足日益严峻的交通需求的主要手段，同时地铁建设产生了大量的深基坑工程,而这些基坑常常处于建筑物密集的市区，施工场地狭小，从而导致基坑工程设计、施工中的问题越来越突出，极大地影响了基坑施工安全和周边建筑物的保护，尤其是不良土质问题给工程安全带来了极大的隐

患。

武汉地区老粘土作为一种膨胀土，胀缩变形是它作为一种不良土质的主要特性。国内外对膨胀土质的各种特性及其土质改良的理论和方法有了比较深入的研究，但对膨胀土质深基坑邻近建筑物的变形研究却较少。本文综合分析了武汉地区老粘土的工程性质及变形机理和基坑开挖对周边环境的影响。此外，通过对沉降监测数据建立统计分析模型，阐述了运用 Elman 神经网络预测模型来处理反馈监测数据的可能性。在此基础上，通过武汉地铁二号线光谷广场站深基坑开挖这一具体的工程实例，介绍了在武汉特有的具有膨胀特性的老粘土地质条件下，基坑开挖所引起的周边建筑物地面出现不同程度的损坏和裂缝的情况，找出建筑物沉降变形影响因素，提出相应的施工技术手段和管理措施，从而达到控制受影响建筑物的沉降和保护建筑物的目的。

最后，本文通过对该工程实践案例控制措施的分析和总结，为今后类似工程地质条件及施工条件的地铁基坑建设安全和建筑物保护提供一定的的借鉴意义。

关键词：  膨胀土   基坑开挖   差异沉降   建筑物变形   Elman 神经网络模型

Abstract

   In recent years, with the rapid development of urbanization, many big cities faced tomany challenges such as traffic and lack of land, the construction of metro became thebest way to solve those problems. At the same time, the construction ofsubway laded to alot of deep foundation pit projectswhich is surrounded with dense buildings in the cities.The  space  of  their  construction  is  narrow.  So  the  design  of  foundation  pit  project  andspecial soil problem in construction ismore and more outstanding. It greatly influencedthe  safety  of  foundation  pit  construction  and  the  protection  of  foundation  pit’s surrounding buildings  

   As a kind of expansive soil,The main characteristic of the old clay in Wuhan districtis  swellshrink  deformation.  All kinds  of  researches  from  domestic  and  foreign  aboutcharacteristics  of  expansive  soil  and  its  soil  improvement theory  method  have  beenthorough. But the studies on the deformation of swelling soil foundation pit’s surroundingbuildings  are  less.  This  paper  analyses  the  engineering  properties  and  deformationmechanism  of  the  old  clay  comprehensively,  and  the  influence  of  foundation  pitexcavation to the surroundings. In additionby the way of settlement monitoring data tobuild  a  statistical  analysis  model,  explaining  the  possibility  that  using  Elman  neuralnetwork forecast model to process and feedback monitoring data. On the base of analysisabove, Taking the deep foundation pit project of optical valley square station of Wuhanmetro  line  2  as  an  example,  introducing  that  damage  and  crack  appear  more  or  lesssurrounding buildings ground under the geological condition of Wuhan district old clayand  influence  of  foundation  pit  excavation.  Then  finding  out  the  factors  of  buildingsettlement and deformation and proposing to take corresponding construction technologyand management measures. At last, the aim of Controlling affected buildings’ settlements and protecting buildings are reached.      In  the  end,  through  the  analysis  and  summary  of  the  control  measures  of  thisengineering  practice  case, theresearch  measures  and  conclusions  can  be  used  asreference for the similar engineering conditions in the constructionof subway foundation pit.

Keywords: expansive soil    foundation excavation    differential settlement building deformation    Elman neural network model