基于K-means聚类方法和Ⅰ Index聚类有效性检验指标的岩体结构面自动分组及应用  

An Auto-classification of Rockmass Discontinuities Based on K-means Cluster Analysis and Cluster Validity Index Ⅰ: Research and Application

作　　者：王俊智 杜朋召 牛兆轩 WANG Jun-zhi;DU Peng-zhao;NIU Zhao-xuan(Department of Geological Engineering,Yellow River Engineering Consulting Co.,Ltd.,Zhengzhou 450003,China;School of Water Resources and Environment,China Universityof Geosciences,Beijing 100083,China)

机构地区：[1]黄河勘测规划设计有限公司地质工程院,郑州450003 [2]中国地质大学(北京)水资源与环境学院,北京100083

出　　处：《长江科学院院报》2018年第9期109-113,120,共6页Journal of Changjiang River Scientific Research Institute

基　　金：黄河勘测规划设计有限公司自主研究开发项目(2009-ky35(2),2017-ky02)。

摘　　要：岩体结构面产状数据的统计分组是工程地质、水文地质工作中基础但十分重要的环节。常用的倾向、走向玫瑰花图和极点等密度图人为主观因素大,需借助合适的数学手段对结构面产状数据进行客观划分。基于Kmeans聚类方法和Ⅰ Index聚类有效性检验指标,提出了一种岩体结构面自动分组方法,并开发了岩体结构面自动分组程序RDAP。通过与经典文献进行对比,验证了所提分组方法的可靠性。最后,以某工程为例,使用RDAP对实测涌水裂隙资料进行了预处理,初步计算了灌浆钻孔的最佳方位,为工程涌水的防治提供了依据。Classifying the orientations of rock discontinuities is a fundamental but critically important routine of engineering geology and hydrogeology.The commonly-used rose diagrams of dip directions and strikes and poles to orientations are subjective.It is appreciated to resort to mathematical approaches.In this paper,an automatic grouping method is proposed based on the algorithms of K-means cluster analysis and cluster validity index I,and a Rock Discontinuities Auto-classification Program(RDAP)is developed.By comparing with Shanley and Mahtab’s result(1976),the reliability of the new grouping method or RDAP is verified.In the end,RDAP is used to cluster the gushing fractures to aid in the selection of the optimum position of grouting drillings of a project,which provides a basis for the prevention and control of water gushing.

关 键 词：岩体结构面 自动分组 K-means聚类方法 IIndex聚类有效性检验指标 RDAP 灌浆钻孔 最佳方位 

分 类 号：TU45[建筑科学—岩土工程]