论文部分内容阅读
概率反分析能够有效地考虑岩土体参数不确定性并融合现场监测数据和观测信息等更新岩土体参数统计特征,进而使得边坡稳定性评价更为符合客观工程实际,然而目前参数概率反分析几乎没有考虑参数固有空间变异性的影响。结合多重响应面和子集模拟提出了考虑岩土体参数空间变异性的边坡参数概率反分析方法,并以芝加哥国会街切坡为例,融合边坡失稳和滑动面入滑点与出滑点的大致位置这两个现场观测信息,概率反分析得到边坡不排水抗剪强度参数的后验统计特征。结果表明:本文提出方法可以有效地解决考虑参数空间变异性的低概率水平边坡参数概率反分析问题,具有较高的计算效率。子集模拟中每层随机样本数目对计算结果具有重要的影响,常用的500组样本点难以获得满意的计算结果。此外,土体参数空间变异性对概率反分析计算结果具有重要的影响,考虑参数空间变异性边坡参数由平稳随机场更新为非平稳随机场,与工程实际相符,然而忽略参数空间变异性更新后的参数仍服从平稳分布。
Probabilistic back analysis can effectively consider the uncertainty of rock and soil parameters and integrate the statistical characteristics of rock and soil parameters such as on-site monitoring data and observation information to make the slope stability evaluation more in line with the objective engineering practice. However, The analysis hardly takes into account the influence of the inherent spatial variability of the parameters. Combined with multi-response surface and sub-set simulation, a parametric inverse slope analysis method is proposed to consider the spatial variability of rock and soil parameters. Taking the cut in Capitol Hill in Chicago as an example, the slope instability and slip-in and slip- The approximate location of the two points of the site observation information, probabilistic back analysis of the slope undrained shear strength parameters of the posterior statistical characteristics. The results show that the proposed method can effectively solve the problem of inverse probability analysis of slope parameters with low probability and considering the spatial variability of parameters, and has high computational efficiency. The number of random samples per sub-set has an important impact on the calculation results, and it is difficult to obtain satisfactory results with the commonly used 500 sets of sample points. In addition, the spatial variability of soil parameters has an important influence on the results of probability back analysis. The slope parameters considering the spatial variability of parameters are updated from stationary random fields to non-stationary random fields, which is consistent with engineering practice. However, the spatial variability of parameters is ignored After the parameters are still subject to a smooth distribution.