【作者】 孙景领；

【导师】 黄腾；

【作者基本信息】 河海大学 ， 大地测量学与测量工程， 2006， 硕士

【摘要】 高速运行的地铁车辆对于轨道的平顺度要求极高,而地铁隧道是承载列车运行的构筑物,其结构体狭长,柔性度大,当结构受力失衡后,极易使狭长的隧道发生局部变形,从而造成轨道线形的变化,影响列车的运行,因此,进行有关隧道结构沉降监测的研究具有重大的现实意义。本文就地铁隧道结构的沉降监测主要研究了以下内容: (1) 总结了影响地铁隧道结构沉降的多种因素;根据地铁隧道本身的特点,论述了沉降监测基准网的布设与施测方法,隧道结构的沉降监测及监测数据的处理与分析,并以南京地铁一号线西延线隧道结构沉降监测为实例,具体阐述了地铁隧道结构沉降监测的过程。 (2) 分析了监测基准网的平差基准及其选取;重点研究了相同基准下监测基准网稳定性分析的几种方法—限差检验法、t检验法、平均间隙法和分块间隙法,并结合南京地铁一号线隧道结构沉降监测基准网的监测数据,对其进行了整体和单点稳定性检验。 (3) 综合论述了灰色系统理论与模型的建立,模型的特征及其数据处理的基本方法;着重讨论了灰色系统理论中的GM(1,1)模型的建模过程及其算法流程。 (4) 阐述了神经网络模型的基本原理,主要包括神经网络的结构、传递函数、学习规则和训练方式;详细讨论了BP神经网络模型的概念、基本结构、神经元模型和算法流程。 (5) 研究了灰色系统理论与神经网络组合的灰色神经网络GNNM(1,1)模型的建模思想、网络结构及其优化GNNM(1,1)模型的方法和学习算法;结合南京地铁一号线隧道结构沉降监测实例,分别运用灰色系统理论GM(1,1)模型和灰色神经网络GNNM(1,1)模型对监测数据进行了模拟与预报。更多还原

【Abstract】 Metro vehicle of quick-speed wheeling extreme highly require rail planeness. And metro tunnel is structure to load-supporting vehicle wheeling, its construction is slender, great flexible degree, when its construction subject to unbalance force, partial slender tunnel easily changes of the form, makes a lot of changes to line shape of rail and affect operation of vehicle. It has grave actual purpose to research sedimentation monitoring of tunnel construction. This paper mainly study as follows contents about sedimentation monitoring of tunnel construction:(1) Generalize multiform factors of influencing settlement of metro tunnel. According to itself features of metro tunnel, discuss means of layout and measuring about settlement monitoring reference net, sedimentation monitoring of tunnel construction and treatment of monitoring data, and giveing one example about settlement monitoring of tunnel construction of Nanjing Metro NO1, specificly expound settlement monitoring of metro tunnel construction.(2) Analyse adjustment reference and itself access of monitoring reference net. Mainly research some analysis methods about stabilization of monitoring reference net based on same reference, including method of inspection of tolerance, t inspection method, average clearance inspection method, blocking clearance inspection method, and based on monitoring data of tunnel construction settlement monitoring reference net of Nanjing Metro NO1, individually use upward methods to inspect stabilization of single-point and global net.(3) Synthetically expound grey system theory and building mode, feature of mode , and treating method of data. Majorly study building process of GM(1,1) mode and algorithm circuit.(4) Represent basic axiom of nerural net, mainly including mode structure, transfer function, learning regulation and training way. Detailed discuss concept, mode, basic fabric, nerve unit mode and algorithm process of BP neural network.(5) Study modelling thought, network configuration, majorize GNNM(1,1) mode method and learning algorithm of GNNM(1,1) mode combined grey system theory and neural network. Based on tunnel construction settlement monitoring of Nanjing Metro NO1, simulate and forecast monitoring data by GM(1,1) mode and GNNM(1,1) mode.更多还原