融合地基SAR与多传感器的深基坑形变监测关键技术及破坏机理研究

Deformation damage of deep foundation pits is a very complex process during the construction management stage, it is often caused by the failure of part support structures or soil instability, and leads to large-scale collapse with serious consequences. This project attempts to integrate both multi-sensor and ground-based SAR data for foundation pit monitoring, to obtain high precision three-dimensional deformation, intelligent inversion and early warning of the occurrence of deep foundation pit deformation and damage. To do this, we need to: study the precision registration of GB-SAR images and atmospheric phase filtering separation technology, correcting the ‘orbital’ relocation errors, extract high coherence point of discontinuous ground-based SAR monitoring, and focus on multi-sources data fusion theory and GB-SAR phase unwrapping based on three-dimensional deformation monitoring data constraints. The establishment of high precision monitoring, simulation, inversion and early warning model would optimize the deformation mechanics and internal structural parameters of the deep foundation pit. In addition to building a network intelligent monitoring and early warning management system about the deep foundation pit, and providing a theoretical basis for safety management and more scientific construction organization. This has an important academic value and practical significance, in promoting deep foundation pit deformation monitoring theory and its development, and in revealing the deformation and damage mechanism of deep foundation pit.

深基坑工程在施工管理阶段的形变破坏是一个非常复杂的过程，往往由局部支护结构破坏或土体失稳引起，导致大范围倒塌，后果非常严重。本课题试图融合基坑监测多传感器与地基SAR数据，获取高精度三维形变，智能反演与预警深基坑形变破坏的发生发展。主要研究GB-SAR影像精密配准和大气相位滤波分离技术，研究非连续地基SAR监测的“轨道”重定位误差改正与高相干点提取，重点研究多源数据融合理论与三维形变监测数据约束的GB-SAR相位解缠，建立高精度深基坑监测、模拟、反演与预警模型，优化基坑形变破坏的力学机制与内部结构参数，构建网络化智能监测预警管理系统，为安全管理与施工组织再优化提供理论基础；对推动深基坑形变监测理论与发展，揭示其形变破坏机理等方面具有重要的学术价值和现实意义。

针对地铁隧道及深基坑形变监测，融合多传感器数据，突破了复杂工况高噪声背景下形变信号提取的关键技术，解决制约三维激光扫描、GPS、地下测斜和地基SAR等测量成果应用于深基坑与隧道地表塌陷自动化精细变形监测的瓶颈问题。形成一套完整的地铁变形监测检测与反演、模拟预测的理论和方法，研制具有自主知识产权的智能监测与预警管控平台，为完善地铁深基坑形变监测体系与失稳灾害预测预防等国家重大需求提供理论与技术支撑。.针对地基SAR特性，提出了一种基于平均相干系数、振幅离差指数等多重阈值优化的PS点选取方法和一种PS大气影响的矫正模型与方法，能有效减弱大气效应的影响。并开发了一种新的实时GBSAR处理方法（RT-GBSAR），提升了地基SAR的形变监测与数据处理能力，分析了重新定位误差的特性和量化效果。将几何和地形改正模型与大气改正模型相结合，利用改进的基于小波变换的信号奇异性检测法，提高了监测区域伪影像检测的效率与准确性。采用大数据量的PS时序处理方法，提高相干点密度，实现地基SAR高精度连续形变监测。结合监测结构特征与动力学特性，建立了地基SAR监测定性与定量评判标准，可实现对构造物形变开裂的早期识别与安全评价。.采用有限元软件的本构理论、库伦模型、摩擦定律对地铁深基坑盖挖逆作法多过程进行数值模拟，基于GIS技术平台组建了可视化三维形变灾害监测预警系统。基于多源形变监测数据，利用 BP 神经网络和数值模拟方法等对隧道与深基坑围岩力学参数进行反演分析，通过离散点拟合来求得隧道与深基坑围岩的自稳系数，为解释地下开挖引起的形变机理奠定了基础。发明了一种带升降式与角度调节的线性扫描地基SAR系统变形监测装置、地基SAR角反射器与申缩观测棱镜等数据配套获取装置，开发了地下土体测斜内外业一体化软件等，实现了多源数据的有效集成费分析与可视化表达。