基于InSAR技术的矿区地表沉降监测及预测.pdf

全日制学术型全日制学术型研究生毕业论文研究生毕业论文 （申请硕士学位）（申请硕士学位） 论 文 题 目论 文 题 目 基于 InSAR 技术的矿区地表 沉降监测及预测 学 位 申 请 人学 位 申 请 人 邓宇声 专 业 领 域专 业 领 域 地理学 研 究 方 向研 究 方 向 地 图 学 与 地 理 信 息 系 统 指 导 教 师指 导 教 师 朱煜峰 2017 年年 6 月月 13 日日 万方数据 Research on Surface Subsidence Monitoring of Coal Mining Area Based on D-InSAR Technology Major Geodesy and Survey Engineering Direction of Study Cartography and Geographic Ination System June ,2017 万方数据 独创性声明独创性声明 本人声明所呈交的学位论文是本人在导师指导下进行的研究工作及 取得的研究成果， 尽我所知， 除了文中特别加以标注和致谢的地方外， 论 文中不包含其他人已经发表或撰写过的研究成果，也不包含本人为获得 其它教育机构的学位或证书而使用过的材料。与我一同工作的同志对本 研究所做的任何贡献均已在论文中作了明确的说明并表示感谢。 作者签名 日期 年 月 日 关于论文使用授权的说明关于论文使用授权的说明 本学位论文作者完全了解东华理工大学有关保留、使用学位论文的 规定东华理工大学有权保留并向国家有关部门或机构送交论文的复印 件和磁盘， 允许论文被查阅和借阅， 可以将学位论文的全部或部分内容编 入有关数据库进行检索， 可以采用影印、 缩印或扫描等复制手段保存、 汇 编学位论文，并且本人电子文档的内容和纸质论文的内容相一致。 保密的学位论文在解密后也遵守此规定。 作者签名 导师签名 日期 年 月 日 论文答辩日期论文答辩日期 年年 月月 日日 万方数据 摘要 I 东华理工大学研究生毕业论文中文摘要首页用纸东华理工大学研究生毕业论文中文摘要首页用纸 毕业论文题目 基于 InSAR 技术的矿区地表沉降监测及预测 地理学 专业 2014 级硕士生姓名 指导教师（姓名、职称） 摘要 在煤矿开采过程中会造成各种环境地质灾害，例如地表塌陷、产生裂缝以及山体 滑坡等，不仅危及人身财产安全，还会对生态 环境造成严重破坏，产生 社会恐慌等不 良 影响，限制了 当地经济的快速发展。煤矿开采区发生的沉降，可能缓慢发生，也可 能发展快速，为了保证安全生产，就得通过一定的技术方法来监测沉降。有一种方法 即合成 孔径雷 达差分干涉测量技术简称 D-InSAR 在近年来 发展迅速，已被证明是 一 种监测地面形变 十分有效率的方法， 能够得到试验区短时间间隔内在空间上连续的地 面形变相关信息，测量精度更可达到毫米级。国内外研 究者开展了 很多利用 D-InSAR 技术监测矿区地表沉降 的相关研究2009 年，Hanssen R 等人 使用该 技术手段对位于 波兰地区的 Wieliczka 盐矿开采区进行地表沉降研究，最终得到了该区域的 地表形变 场数据；2012 年， 在位于美国西部的犹他州 Book Cliffs 煤矿区域，James Donovan 等 人使用 D-InSAR 方法进行了相关研究， 得到该试验区域 的三维地表形变图； 国内研究 者任小冲等根据 ALOS 卫星获取的青藏高原 冻土区域影像数据，结合 D-InSAR 技术 得到了该地区的形变图。因为时空失相干 及 大气效应等因素会 影响到 D-InSAR 的监 测结果，为了能够 有效解决这些影响，近年来 又发展出 永久散射体（PS） 、小基线集 （SBAS） 、 最小二乘LS 等基于 时间序列的形变分析方法，对于监测较长时间内大范 围的地表的缓慢形变这些方法非常有效， 能较快速分析试验区域的演化规律以及沉降 发展情况。相较于以往的研究多集中在相干性好的平坦地区，本文选择的试验区位于 江西省的西部，属丘陵地势，深山密林，植被丰茂，受季节性影响，对于 D-InSAR 的 研究具有一定的地域针对性，可实现一定的研究价值。 本文以江西省丰城某煤矿区为例， 利用 D-InSAR 技术和小基线技术 SBAS, Small Baseline Subset对地表沉降监测的理论和方法进行了探讨；基于半参数灰色模型，利 用补偿最小二乘准则下的 GM（1,1）模型结合 D-InSAR 技术监测数据模拟计算了 煤 矿区地表沉降机理。主要研究内容包括 1叙述了 D-InSAR 技术的基本理论 与 实现方法，讨论了 D-InSAR 技术的误差来 源。 2根据获取 的 PALSAR 雷达影像数据，基于瑞士 GAMMA 公司开发的 D-InSAR 技术处理软件 GAMMA 软件，详细介绍了 “二轨”差分法数据处理流程 和相关命 万方数据 东华理工大学硕士学位论文 II 令， 并针对二轨差分方法平地相位去除算法对精度的影响，分别讨论了不同算法带来 的误差源。利用长波段 PALSAR 雷达影像数据，结合 “二轨”差分法得到矿区地表形 变场， 并且使用水准数据来验证 D-InSAR 技术的监测结果的准确性。 3为了解决常规 D-InSAR 技术中时间分辨率差和高程 DEM, Digital Elevation Model误差带来的问题，利用 高级 D-InSAR 技术小基线集 SBAS 技术得到实验矿 区地表的 时间序列沉降图 以及累积形变量。 4 基于半参数灰色模型，分别利用 GM（1,1） 和半参数 GM（1,1）两种 模型结 合 D-InSAR 技术监测数据 进行拟合， 然后分析两种模型的 拟合精度， 最后采用 半参数 GM（1,1）模型进行预测。一方面能够打破 传统平差 模型存在的局限性，另一方面能 使采用的 数学模型与客观实际 更加接近，在数值 上可以分别计算出数学模型的误差 （非参数分量） 和偶然误差的 估值，能够更加充分的利用观测值中所获取的信息，模 拟计算了 煤矿区地表沉降机理。 本文通过对江西省 丰城煤矿 开采区沉降监测 进行了 实验研究，验证了 适用于煤 矿区常规的 D-InSAR 技术以及 SBAS 方法监测的准确性，对地质条件与气候都具有地 域性的南方煤矿区开采的雷达遥感沉降监测具有十分重要的意义。最后，基于半参数 灰色模型，结合 D-InSAR 获得的监测数据对矿区沉降进行了拟合，通过对预测数据 的分析了解矿区沉降规律以及下阶段沉降趋势，为矿区预防重大灾害，保证安全生产 提供了重要的判断依据。 关键词关键词D- InSAR，地表沉 降监测，小基 线集，半参数 灰色模型，模拟 计算 万方数据 ABSTRACT III 东华理工大学研究生毕业论文英文摘要首页用纸东华理工大学研究生毕业论文英文摘要首页用纸 THESISMonitoring and Forecasting of Surface Subsidence in Mining Area Based on InSAR Technology SPECIALIZATION Geodesy and Survey Engineering POSTGRADUATE MENTOR ABSTRACT In the process of coal mining will cause a variety of environmental geological disasters, such as surface subsidence, cracks and landslides, not only endanger the safety of personal property, but also serious damage to the ecological environment, resulting in social panic and other adverse effects, limiting the local economy the rapid development. Coal mining area occurred in the settlement, may occur slowly, may also develop rapidly, in order to ensure safe production, you have to through certain technical means to monitor the settlement. Synthetic Aperture Radar Differential Interferometry D-InSAR has been developed rapidly in recent years and has been proved to be a very efficient for monitoring ground deation. It is possible to obtain continuous ground deation related ination within the short time interval of the test area, measurement accuracy can reach the millimeter level . Researchers at home and abroad have conducted a number of studies on the use of D-InSAR technology to monitor surface subsidence in the mining area. In 2009, Hanssen R et al. Used D-InSAR techniques to study the subsidence of the Wieliczka salt mining area in Poland Area of the surface deation data; In 2012, James Donovan et al. Conducted a study using the D-InSAR in the Book Cliffs coal mine area of Utah in the western United States, and obtained the three-dimensional surface deation map of t Based on the acquired PALSAR radar image data, the data processing flow and related commands of the “two-track“ difference are introduced in detail based on the D-InSAR technology processing software GAMMA developed by GAMMA Company in Switzerland, and the two-track differential is used to remove the ground The effects of different algorithms on the error are discussed.he test area; The domestic researcher Ren Xiaocheng and other use of PALSAR satellite L-band radar data, the Qinghai-Tibet Plateau area for differential treatment, access to the region of permafrost deation map, and demonstrated the D-InSAR in the Qinghai-Tibet Plateau permafrost area to carry out the possibility of deation monitoring. Because the space- time incoherence and atmospheric effects and other factors will affect the D-InSAR monitoring results ,In order to be able to effectively solve these effects, in recent years, the development of permanent scatterers PS, small baseline set SBAS, least squares LS, and so on other time-based deation analysis s, for monitoring a long time on a 万方数据 东华理工大学硕士学位论文 IV wide range of surface rough deation of these s is very effective, can quickly analyze the evolution of the experimental area and settlement development. Compared with the previous studies, the experimental area is located in the western part of Jiangxi Province, which belongs to hilly terrain, mountain dense forest, rich vegetation and seasonal effect. It has certain characteristics for D-InSAR research. Of the geographical relevance, can achi Based on the acquired PALSAR radar image data, the data processing flow and related commands of the “two-track“ difference are introduced in detail based on the D- InSAR technology processing software GAMMA developed by GAMMA Company in Switzerland, and the two-track differential is used to remove the ground The effects of different algorithms on the error are discussed.eve a certain value of research. In this paper, a coal mine in Fengping of Jiangxi Province, for example, the theory and of surface subsidence monitoring are discussed by using D-InSAR technology and Small Baseline Subset; Based on the semi - parametric gray model, the GM 1,1 model under the compensation least squares criterion is used to simulate the ground subsidence mechanism of coal mining area by D - InSAR technology. The main research contents include 1 The basic theory and implementation of D-InSAR technology are briefly described, and the error source of D-InSAR technology is briefly discussed. 2 Based on the acquired PALSAR radar image data, the data processing flow and related commands of the “two-track“ difference are introduced in detail based on the D-InSAR technology processing software GAMMA developed by GAMMA Company in Switzerland, and the two-track differential is used to remove the ground The effects of different algorithms on the error are discussed.The PALSAR radar image data of long band is combined with the “two-track“ differential to obtain the surface deation field of the mining area, and the level data are used to verify D -InSAR technology to monitor the accuracy of the results. 3 In order to solve the problem of error in time difference and elevation DEM in conventional D-InSAR technology, the time series subsidence map of experimental surface is obtained by using advanced D-InSAR technique - small baseline set SBAS As well as cumulative deation. 4 Based on the semi-parametric gray model, GM 1,1 and semi-parametric GM 1,1 model were used to fit the D-InSAR technique to fit the data. Then analyzed the fitting precision of the two models. Semi-parameter GM 1,1 model for prediction. On the one hand, it can break the limitation of the traditional adjustment model, on the other hand, make the mathematical model and the objective reality closer together, and calculate the error of the mathematical model nonparametric component and the accidental error , The mechanism of ground subsidence in coal mining area can be calculated and calculated by using the ination obtained in the observation value more fully. In this paper, the experimental study on the settlement monitoring of the Fengcheng coal mining area in Jiangxi Province,which carried out to verify the accuracy of the conventional D-InSAR technology and SBAS for coal mining area. The geological conditions and climate are regional southern coal mining area Mining radar remote sensing settlement monitoring is of great significance. Finally, based on the semi-parametric gray model and the monitoring data obtained by D-InSAR, the settlement of the mining area is 万方数据 ABSTRACT V fitted. The settlement rule of the mining area and the settlement trend of the next stage are analyzed by analyzing the forecast data. To provide a reference for monitoring the surface subsidence,and it is a great significance to prevent the collapse of the mine. Key WordsD-InSAR, surface subsidence monitoring, small baseline set, semi - parametric gray model, simulation calculation 万方数据 目录 V 目录 摘要摘要 ........................................................................................................................................ I Abstract ............................................................................................................................. III 表目录表目录 ............................................................................................................................... VII 图目录图目录 ............................................................................................................................... VII 第第 1 章章 绪论绪论 ...................................................................................................................... 1 1.1 研究背景与意义 .................................................................................................. 1 1.2 国内外研究现状 .................................................................................................. 2 1.2.1 D-InSAR 技术的发展 ............................................................................... 2 1.2.2 D-InSAR 技术在矿区沉降监测的应用现状 ........................................... 4 1.3 研究目的 .............................................................................................................. 4 1.4 论文章节安排 ...................................................................................................... 5 第第 2 章章 D-InSAR 技术的基本理论技术的基本理论................................................................................. 7 2.1 InSAR 技术.......................................................................................................... 7 2.1.1 InSAR 技术基本原理 ............................................................................... 7 2.1.2 InSAR 提取 DEM 数据处理流程 ............................................................ 9 2.2 D-InSAR 技术 ................................................................................................... 10 2.2.1 D-InSAR 技术基本原理 ......................................................................... 10 2.2.2 D-InSAR 监测方法 ................................................................................. 11 2.3 影响监测精度的因素 ........................................................................................ 12 2.3.1 InSAR 平地相位去除算法和对 DEM 精度影响分析 .......................... 16 2.3.2 基于轨道数据算法 ................................................................................. 16 2.3.3 基于干涉频谱算法 ................................................................................. 17 2.3.4 基于轨道参数算法的误差源分析 ......................................................... 18 2.3.5 基于干涉频谱算法的误差源分析 ......................................................... 19 2.4 D-InSAR 技术数据处理实现过程 ................................................................... 20 2.4.1 实验数据与处理软件 ............................................................................. 20 2.4.2 ALOS-PALSAR 卫星数据介绍 ................................................................ 20 2.4.2 SRTM DEM ............................................................................................. 21 2.4.3 处理软件介绍 ......................................................................................... 21 2.5 D-InSAR 二轨差分技术的处理流程 .............................................................. 21 2.6 实例数据分析 .................................................................................................... 27 2.6.1 干涉对的选择 ......................................................................................... 27 2.6.2 差分干涉图的生成及滤波 ..................................................................... 30 2.6.3 相位解缠以及沉降图的生成 ................................................................. 30 2.7 本章小结 ............................................................................................................ 31 第第 3 章章 基于基于 SBAS 技术的矿区地表沉降监测实例技术的矿区地表沉降监测实例 ................................................... 33 3.1 引言 .................................................................................................................... 33 3.2 小基线集SBAS技术原理 ............................................................................... 33 3.2.1 最小二乘方法LS .................................................................................. 33 3.2.2 奇异值分解SVD .................................................................................. 35 万方数据 东华理工大学硕士学位论文 VI 3.2.3 线性形变速率和高程误差的的提取 ..................................................... 35 3.3 实验研究 ........................................................................................................... 36 3.3.1 研究区域概况 ......................................................................................... 36 3.3.2 数据选择 ................................................................................................. 37 3.3.3 数据预处理 ............................................................................................. 38 3.3.4 高相干点的选择 ..................................................................................... 40 3.3.5 时间序列沉降图的生成 ......................................................................... 41 3.4 结