【作者】 陈珂；

【导师】 张献州；

【作者基本信息】 西南交通大学 ， 大地测量学与测量工程， 2013， 硕士

【摘要】 伴随着信息化测绘时代的来临,以数字化信息为核心的传感网技术对工程领域中的设计、施工、监测和管理模式产生了深远影响。考虑到隧道工程的变形监测内容多、周期长、数据量大等特征,以前纯人工定时实地的监测方法难以满足实际需求,借助卫星定位、测量机器人、传感器、网络通讯等技术,传感网模式的现代化监测系统相继产生。新疆盐水沟隧道是西气东输的示范工程,由于地质复杂、环境恶劣,隧道内部出现多处不同程度的开裂,对内部管道有较大的安全隐患。为了确保管道运输线路的安全,本文基于传感器网络技术,引入传感器优化设计和组合预测模型,设计研发了集工程前期优化设计、数据采集、通讯、计算、预测为一体的变形监测传感网预警原型系统。本文共分五章：1.第一章介绍传感器、传感网和监测传感网的发展现状以及常见的传感网监测预警系统,并针对本文研究内容分析了国内外系统的优缺点和适用性。2.第二章研究了传感器的数目、定位和布线优化模型,并根据盐水沟隧道实际情况,运用动态规划原理,以线缆价格、长度和接线头个数为制约因子分析传感器布线方案中最优方案的选择。3.第三章详细阐述了传感网模式下隧道变形监测特点,在传感网技术架构基础上,系统总结了传感网的组成部分以及实施流程。4.第四章以实际工程为背景,建立了灰色理论、时间序列和BP神经网络三种单项预测模型,通过对实测及仿真数据的分析与研究,探讨建立组合预测模型的必要性,然后对比分析可变加权系数和最优加权系数2种组合预测模型预测精度,并给出了两种方法的适用条件。5.第五章设计并实现了隧道变形监测传感网预警系统的原型,并以盐水沟隧道监测项目为平台,对系统部分功能做了测试与应用。更多还原

【Abstract】 With the arrival of the informationization surveying and mapping era, sensor technology taking the digital information as the core, has had a far-reaching influence on the original optimal design, monitoring and management. Considering the deformation of tunnel engineering has the characteristics such as monitoring content more, long cycle, large amount of data, the monitor method of traditional periodic, periodic artificial is difficult to meet the practical demand. The modern monitoring systems based on the sensor network model have being emerging, which are supported by the techniques such as satellite positioning, surveying robot, sensor network and network communication technologies.The Yanshuigou tunnel is the key of the West-East natural gas transmission project. Due to the complex geology and harsh environment of the tunnel site, the internal structure of the tunnel have appear several cracks, which have some potential safety hazard to the pipes. Based on sensor network technology, this paper introduces optimal design and combination forecasting model then designs a monitoring prototype system, which contains route optimization, data acquisition, processing, management, analysis and prediction functions. This thesis is composed of five chapters.1. The first chapter introduces the development of sensors, sensor network and the network for monitoring and the frequently used monitoring system based on sensor network mode. At the same time, this chapter analyzes the advantages and shortcomings and the application situation of systems inside and outside.2. The second chapter studies the optimization models of sensor which includ the number, location and routing. It also use the dynamic programming principle to develop the liner optimal program according to actual situation of the Yanshuigou tunnel.3. The third chapter sets forth the characteristics of deformation monitoring in tunnel. Based on sensor network mode, the components and implement process of sensor network is summarized systematically.4. In the forth chapter, regarding actual engineering as the back, three prediction models of grey theory, time series and BP neural network are established. Based on analysis and comprehension of the measured and simulation data, this chapter explores the necessity of building the combining forecast model. The applicable conditions of two methods are presented by comparing the prediction accuracy of the variable weight and fixed weight combination model. 5. Through above research, a sensor network early warning prototype system for deformation monitoring was designed and developed. Meanwhile, system uses the Yanshuigou tunnel monitoring project as a platform to test its partial function.更多还原