溶洞上覆土层触探特征分析

发布时间：2018-02-24 20:08

  本文关键词： 岩溶 软弱土层 土洞 概率　出处：《综合运输》2016年12期 　论文类型：期刊论文

【摘要】：本文首先介绍基于归一化的锥尖阻力Q_t和归一化的摩阻比F_R等触探参数的Robertson分类图及其解析表达形式,再根据研究区域的土层分布状况,简化Robertson分类图和贝叶斯模型。最后收集江西省某高速公路沿线溶洞上覆土层89个CPT和钻孔取样资料,分别采用最大似然法和贝叶斯法对Robertson分类图中的边界进行修正和比较,发现变异系数COV越小,先验分布越接近Robertson土壤分类图,预测结果越接近于Robertson土壤分类图,准确率也较以往的70%有所提高。变异系数COV越大,先验分布越为含糊不清,预测结果越接近于最大似然法的结果,但由于考虑了先验分布,准确率仍高于最大似然法。而在样本数量有限的情况下,最大似然法计算结果与Robertson分类图存在较大差别,准确率较差,应谨慎使用。
[Abstract]:This paper first introduces the Robertson classification diagram of penetration parameters such as normalized cone tip resistance Q _ t and normalized friction ratio _ F _ R, and its analytical expression, and then according to the distribution of soil layer in the study area, The Robertson classification map and Bayesian model are simplified. Finally, 89 CPT and borehole sampling data are collected from the overlying soil layer of a karst cave along a highway in Jiangxi Province, and the boundary of the Robertson classification map is revised and compared by using the maximum likelihood method and Bayesian method, respectively. It was found that the smaller the coefficient of variation (COV) was, the closer the prior distribution was to the Robertson soil classification map, the closer the predicted result was to the Robertson soil classification map, and the higher the accuracy was compared with the previous 70%. The larger the coefficient of variation COV was, the more ambiguous the prior distribution was. The prediction results are closer to those of the maximum likelihood method, but the accuracy is still higher than that of the maximum likelihood method because the prior distribution is taken into account. In the case of limited number of samples, the results of the maximum likelihood method are quite different from that of the Robertson classification map. The accuracy is poor, should be used carefully.
【作者单位】： 江西交通咨询公司;江西省交通科学研究院;
【基金】：国家自然科学基金(51508246) 交通运输部重点科技项目(2013318780290) 江西省交通运输厅科技项目(2015C0022)
【分类号】：U412.22
，

本文编号：1531545