System of Functional Diagnosis for Stock Management of Soil Structures Based on Integrated Measurement/Data Assimilation

本研究旨在建立集测量、数据同化、风险评估和LCC估算于一体的土壤结构物库存管理功能诊断系统。讨论了老化土坝强度参数的空间分布及其识别方法。通常情况下，填土的强度是由SPT N值来预测的。而在本研究中，采用瑞典重量发声测验(SWS测验)来获得N值的空间分布，这是一种更简单的方法，并根据发声测验确定了N值的统计模型。为此，采用了最小化Akaike‘s Information Criteria(MAIC)方法和半方差函数方法来识别空间相关性特征。N值的空间分布是由高分辨率的测深试验确定的，因为点估计是用Short in…获得的更多的蚊子。为了对点估计值进行内插，采用了地统计学方法中的指示器模拟法。在该方法中，可以同时使用硬数据(主数据)和软数据(补充数据)。SWS的结果和地球物理勘探方法中的面波方法(SWM)的结果分别作为硬数据和软数据处理。综合这两个结果，可以识别出N值的准确空间分布。提出了一种将有限元模型与基于粒子滤波的数据同化相结合的计算方法，用于根据沉降和孔压的实测数据识别土结构内部的劣化区。在基于可靠性的设计框架下，对路堤加固效果与路堤抗震、暴雨安全性进行了评价。土结构的风险和LCC是根据上面列出的总体考虑进行评估的。较少

This research aimed to establish the system of functional diagnosis for stock management of soil structures using integrated measurements, data assimilation, risk evaluation, and estimation of LCC. The spatial distribution of the strength parameters of decrepit earth-fill dams, and the identification methods of distribution are discussed. Generally, the strength of the earth-fill is predicted from the SPT N-values. While, in this research, the Swedish Weight Sounding Tests (SWS Tests) are conducted to obtain the spatial distribution of the N-values as the simpler method, and the statistical model of the N-values is determined based on the sounding tests. For this task, the Minimizing Akaike's Information Criteria (MAIC) method is employed, and the semi-variogram method is also used to identify the spatial correlation characteristics. The spatial distribution of the N-value is identified from the sounding tests with high resolution, since the point estimations are obtained with short in … More tervals. To interpolate the point estimate values, the indicator simulation method, which is one the geostatistical methods, is employed. In the method, the hard data (primary data) and the soft data (complementary data) can be used simultaneously. Results from the SWS and the Surface Wave Method (SWM), which is one of the geophysical exploration methods, are dealt with as hard and soft data, respectively. With synthesizing two results, the accurate spatial distribution of the N-values can be identified. A computational method, incorporating the finite element model into data assimilation using the particle filter, is presented for identifying deteriorated area within soil structures based on measured data on the settlement and the pore pressure. The effect of improving of the embankment is evaluated in relation to the safety of the embankment against the earthquake and the heavy rain within the framework of the reliability-based design. Risk and the LCC of the soil structures are evaluated from the overall considerations listed above. Less