论文部分内容阅读
采用孔压静力触探(CPTU)测试资料进行土的工程分类时,在土成分和力学性状等相关性方面会产生很多不确定性。根据已有的CPTU土分类系统划分土类时,这种不确定性经常会导致不同土类的重叠。聚类分析是一个对相似数据进行分组归类的数学统计方法,基于CPTU测试资料将其应用于地质土层的划分。根据5条高速公路场地的现场CPTU试验和钻孔柱状图资料,阐述了已有聚类分析方法在划分土层界面的应用。结果表明:聚类分析方法不仅能够描绘出土层中主要的变化,而且能够探测到薄土夹层。因此,基于聚类分析理论的CPTU方法可以可靠有效地提供地质土层的初步划分。
When using CPTU test data to classify soil engineering, there is a lot of uncertainty in the correlation between soil composition and mechanical properties. When the soil is classified according to the existing CPTU soil classification system, this uncertainty often leads to overlaps between different soil types. Clustering analysis is a mathematical statistical method for grouping similar data, which is applied to the division of geological soil based on CPTU test data. According to the on-site CPTU test and drilling histogram data of five freeway sites, the application of the existing cluster analysis method in dividing the soil layer interface is expounded. The results show that the clustering method can not only describe the main changes in the soil layer, but also detect the thin interlayer. Therefore, the CPTU method based on cluster analysis theory can provide the preliminary division of geological soil layer reliably and effectively.