基于系统聚类算法的CPT土体分层试验研究  

Experimental Study on CPT Soil Mass Stratification Based on System Clustering Algorithm

作　　者：肖乾 袁颖 汪婷 XIAO Qian;YUAN Ying;WANG Ting(Public Teaching Department,Anhui Technical College of Industry and Economy,Hefei Anhui 230051;School of Urban Geology and Engineering,Hebei GEO University,Shijiazhuang Hebei 050031)

机构地区：[1]安徽工业经济职业技术学院公共教学部,安徽合肥230051 [2]河北地质大学城市地质与工程学院,河北石家庄050031

出　　处：《湖北理工学院学报》2024年第6期36-41,共6页Journal of Hubei Polytechnic University

基　　金：安徽省高等学校自然科学项目(项目编号:2021ZK02);安徽省高等学校哲学社会科学项目(项目编号:2022AH052659)。

摘　　要：为了提高工程勘察中分辨土类及确定土层位置的高效性与准确性,开展了基于系统聚类算法的CPT土体分层试验研究。通过选择变量、标准化变量、欧式距离计算、聚类数目确定及命名、在轨道交通工程中展开试验,阐述了系统聚类算法在CPT土体分层中的应用。模拟划分的结果与实际钻孔柱状图结果对比,具有较高的相似性,说明以q_(c)-F_(R)为参数的聚类分层图能够准确识别土体主层以及次要层的具体位置,也能够将不同深度的相同土体准确地归为一类,在相对复杂的地层中更能够识别出次要层、过渡带、薄夹层等,可以用于地质土层的初步划分。In order to improve the efficiency and accuracy of soil resolution and soil location in engineering survey,the CPT soil stratification test based on system clustering algorithm was conducted.Through variables selection,variables standardization,Euclidean distance calculation,cluster number determination and naming,conducting experiments in rail transit engineering,the application of system clustering algorithm in CPT soil stratification are expounded.The simulated division results and the actual drilling bar chart are compared,and they have high similarity,indicating the clustering diagram with q_(c)-F_(R)parameters can accurately identify the location of soil main and secondary layer,and can also accurately classify the same soil at different depths into one category.In relatively complex strata,it can better identify secondary layers,transition zones,thin interlayers,etc.,which can be used for preliminary division of geological soil layers.

关 键 词：静力触探试验 系统聚类 土层划分 

分 类 号：TU42[建筑科学—岩土工程] P642.11[建筑科学—土工工程]