DInSAR技术在矿区地表重复采动开采沉陷监测中的应用研究

发布时间：2018-03-21 04:07

本文选题：DInSAR　切入点：开采沉陷监测　出处：《中国矿业大学》2014年博士论文　论文类型：学位论文

【摘要】：煤炭开采引起的地表沉降现象十分普遍，加强对矿区地面沉降现象的研究已成为解决四矿问题和实现矿区可持续发展的重要课题之一。对矿区地面沉降进行持续有效监测，可及时掌握沉降规律和地表破坏程度，为合理开采地下矿产资源，控制地面沉降提供决策依据。传统地表沉降监测方法和技术覆盖面小、作业强度大、效率低、不能适应生产发展的需要。差分合成孔径雷达干涉测量（DInSAR）技术具有全天候、全天时、高分辨率和连续空间覆盖的优势，已发展成为新型监测矿区开采沉陷的技术。本文针对DInSAR技术在矿区沉陷监测应用中存在的一些问题进行了较为深入的研究，包括DInSAR监测地表形变的影响因素、基于外部DEM和角反射器的山区高分辨率SAR影像地理编码和时序DInSAR技术及其在矿区地表重复采动开采沉陷监测中的应用等。取得的主要研究成果如下：（1）从理论上分析了基线估计误差和外部DEM高程误差对二轨DInSAR技术监测地表形变的影响，，提出了一种基于地面高相干点的基线估计方法和基于滤波掩模的估计DInSAR干涉图中大气相位的方法，有效补偿或去除了基线估计误差和大气延迟相位的影响。三种波段的SAR影像实证研究表明：利用TerraSAR-X影像监测到13.3cm/11d的形变速率，利用ALOS-PALSAR影像监测到40.6cm/46d的形变速率；相比较而言，EnviSAT-ASAR影像受波长和时空分辨率的限制，监测大形变的能力最弱。（2）研究了不同多视系数和外部DEM对DInSAR技术监测地表沉陷的影响，发现：对于EnviSAT-ASAR和ALOS-PALSAR数据，当外部DEM相同时，基于不同多视系数的干涉处理对形变监测结果在形变量上的影响并不明显；对于高分辨率TerraSAR-X影像，多视相干处理会降低DInSAR监测到的最大形变量，应尽可能选择较小的多视系数。就外部DEM质量而言，对于EnviSAT-ASAR和ALOS-PALSAR数据，其影响主要体现为少量的残余地形相位；对于高分辨率TerraSAR-X影像，SRTM引起的残余相位明显，尤其在地表沉陷区域边界和非沉陷区，SRTM补偿地形相位的能力与Relief-DEM相比，差异显著。（3）提出了一种利用角反射器和模拟SAR影像对山区高分辨率SAR影像进行地理编码和精度评定的方法。自主完成了研究区角反射器的设计、加工与安装，提出了基于点目标响应的提取角反射器在多视强度图上精确位置的方法，在此基础上，将角反射器用于地理编码优化，有效提高了基于模拟SAR影像的地理编码精度。（4）利用X波段SAR数据对微小形变较敏感的优势，综合GPS观测资料、时序DInSAR及不同多视处理的DInSAR监测结果，结合矿区工作面地质采矿资料，成功提取了西山矿区重复采动的地表沉陷边界及重复采动过程中开采沉陷相关参数，如边界角、超前影响角、起动距、下沉系数，实现了DInSAR技术在矿区开采沉陷监测中的定量化应用，为认识重复采动影响下地表沉陷规律提供了新的有效手段。
[Abstract]:The phenomenon of surface subsidence caused by coal mining is very common. It has become one of the important subjects to solve the problems of the four mining areas and to realize the sustainable development of the mining areas to strengthen the research on the phenomenon of the land subsidence in the mining areas. The ground subsidence in the mining areas is continuously and effectively monitored. It can grasp the law of subsidence and the degree of surface damage in time, and provide a decision basis for rational exploitation of underground mineral resources and control of land subsidence. The traditional monitoring methods and techniques of surface subsidence have small coverage, large working intensity and low efficiency. The differential synthetic Aperture Radar Interferometry (DInSAR) technology has the advantages of all-weather, all-day, high-resolution and continuous space coverage. It has been developed into a new technology for monitoring mining subsidence in mining area. In this paper, some problems existing in the application of DInSAR technology in mining subsidence monitoring are studied deeply, including the influencing factors of surface deformation monitoring by DInSAR. Based on external DEM and angle reflector, high resolution SAR image geographic coding and sequential DInSAR technology in mountainous area and its application in mining subsidence monitoring of surface repeated mining are presented. The main research results are as follows:. 1) the effects of baseline estimation error and external DEM height error on surface deformation monitoring by the second track DInSAR technique are analyzed theoretically. A baseline estimation method based on high coherence points on the ground and a method based on filtering mask to estimate the atmospheric phase in DInSAR interferogram are proposed. The effect of baseline estimation error and atmospheric delay phase is effectively compensated or eliminated. The empirical study of SAR images in three bands shows that the deformation rate of 13.3 cm / 11 d is monitored by TerraSAR-X image and 40.6 cm / 46 d by ALOS-PALSAR image. In contrast, EnviSAT-ASAR images have the weakest ability to monitor large deformation due to the limitation of wavelength and space-time resolution. (2) the effects of different multi-view coefficients and external DEM on surface subsidence monitoring by DInSAR technology are studied. It is found that for EnviSAT-ASAR and ALOS-PALSAR data, when the external DEM is the same, Interference processing based on different multi-view coefficients has no obvious effect on the shape variables of deformation monitoring results. For high-resolution TerraSAR-X images, multi-view coherent processing can reduce the maximum shape variables monitored by DInSAR. In terms of external DEM quality, for EnviSAT-ASAR and ALOS-PALSAR data, the influence is mainly reflected in a small amount of residual topographic phase, while for high-resolution TerraSAR-X images, the residual phase is obvious. In particular, the ability of SRTM to compensate the topographic phase in the boundary and non-subsidence areas of surface subsidence is significantly different from that of Relief-DEM. In this paper, a method of geocoding and accuracy evaluation of high resolution SAR images in mountainous area by using angle reflectors and simulated SAR images is presented. The design, processing and installation of corner reflectors in the study area are completed independently. A method of extracting the precise position of corner reflector on multi-intensity map based on point target response is proposed. On this basis, the angle reflector is used to optimize geographical coding, which effectively improves the precision of geographical coding based on analog SAR image. Using the advantage of X-band SAR data sensitive to small deformation, synthesizing GPS observation data, sequential DInSAR and different multi-view processing DInSAR monitoring results, combined with mining face geological and mining data. The surface subsidence boundary of repeated mining in Xishan mining area and the parameters related to mining subsidence during repeated mining are successfully extracted, such as boundary angle, leading influence angle, starting distance, subsidence coefficient, etc. The quantitative application of DInSAR technology in mining subsidence monitoring in mining area is realized, which provides a new effective means to understand the law of surface subsidence under the influence of repeated mining.
【学位授予单位】：中国矿业大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：P642.26;P225

【参考文献】