Next-Generation Imaging Technique for Underground Object Detection

该项目的目的是开发一种先进的计算和实验基础，以便使用弹性(即地震)波对地下物体进行便捷的3D成像。与现有的需要对地下区域进行全体积离散的三维地震分析不同，所提出的成像技术围绕弹性动力学边界积分方程(BIE)方法，从而只渲染隐藏对象的轮廓。这种仅限边界的传感方法提供了巨大的计算节省，其起源于雷达和声纳技术，但在地震调查方面基本上没有人探索过。为了弥补这一差距，本研究的第一阶段致力于提升BIE成像技术的原型，以增加其稳健性，考虑更广泛的地下条件，并处理实验和建模误差的不利影响。该项目的实验阶段旨在将新的传感技术应用到一个紧凑的测试装置中，其中包括一个振动震源和一个表面三轴运动传感器阵列。研究工作包括1)总结声学和结构形状优化方面的最新发现，以开发一种快速地面探测技术，从而提供高保真BIE成像所需的可靠初始猜测；2)扩展该方法以处理受噪声污染的测量和多层地质沉积物；3)综合计算开发，以实现稳健而便捷的三维地震传感方法；4)小规模实施该方法以识别埋在沙仓中的障碍物，以及5)旨在对现有的校园地下设施进行成像的现场工作。在上述研究计划的补充下，将继续努力促进在土木工程中越来越多地使用无损检测。新的传感技术的潜在应用包括快速探测地下设施和未爆弹药；以符合成本效益的方式绘制城市地区地下基础设施的地图；以及非侵入性地划定埋在地下的废物。在这些地区和其他地区，对快速和经济的三维地震成像的需求在电磁勘探可能因其分辨率和穿透深度有限而不充分的情况下尤其明显。

The aim of this project is to develop an advanced computational and experimental basis for expedient 3D imaging of subterranean objects using elastic (i.e. seismic) waves. In contrast to existing three-dimensional seismic analyses that require full volume discretization of the underground domain, the proposed imaging technology revolves around an elastodynamic Boundary Integral Equation (BIE) method so that only the outline of a hidden object is rendered. This boundary-only sensing approach, which offers formidable computational savings, has its origins in radar and sonar technologies, but has been largely unexplored in the context of seismic surveys. Aimed at bridging such gap, the first stage of this research is focused on elevating the prototype BIE imaging technique as to increase its robustness, account for a wider class of subterranean conditions, and deal with the detrimental effects of experimental and modeling errors. The experimental phase of this project aims to implement the new sensing technology into a compact testing setup that involves a vibratory seismic source and a surface array of triaxial motion sensors. The research efforts include 1) generalization of the recent findings in acoustics and structural shape optimization to develop a rapid ground-probing technique and thus furnish a reliable initial guess required for the high-fidelity BIE imaging; 2) extensions of the method to deal with noise-polluted measurements and multi-layered geological deposits; 3) synthesis of the computational developments for a robust, yet expedient approach to 3D seismic sensing; 4) small-scale implementation of the method to identify obstacles buried in a sand bin, and 5) a field effort aimed at imaging an existing, on-campus underground facility. The foregoing research program will be complemented with a continuing effort to foster the growing use of non-destructive testing in civil engineering. Potential applications of the new sensing technology include rapid detection of underground facilities and unexploded ordnances; cost-effective mapping of subterranean infrastructure in urban areas, and non-invasive delineation of buried waste. The need for fast and economic 3D seismic imaging in these and other areas is especially pronounced in situations where electromagnetic surveys may be inadequate due to their limited resolution and depth of penetration