• 全国中文核心期刊
  • 中国科技核心期刊
  • 美国工程索引(EI)收录期刊
  • Scopus数据库收录期刊
王俊杰, 冯登, 柴贺军, 刘云飞. 基于赤平极射投影和K-均值聚类算法的优势结构面分析[J]. 岩土工程学报, 2018, 40(1): 74-81. DOI: 10.11779/CJGE201801006
引用本文: 王俊杰, 冯登, 柴贺军, 刘云飞. 基于赤平极射投影和K-均值聚类算法的优势结构面分析[J]. 岩土工程学报, 2018, 40(1): 74-81. DOI: 10.11779/CJGE201801006
WANG Jun-jie, FENG Deng, CHAI He-jun, LIU Yun-fei. Dominant discontinuities based on stereographic projection and K-means clustering algorithm[J]. Chinese Journal of Geotechnical Engineering, 2018, 40(1): 74-81. DOI: 10.11779/CJGE201801006
Citation: WANG Jun-jie, FENG Deng, CHAI He-jun, LIU Yun-fei. Dominant discontinuities based on stereographic projection and K-means clustering algorithm[J]. Chinese Journal of Geotechnical Engineering, 2018, 40(1): 74-81. DOI: 10.11779/CJGE201801006

基于赤平极射投影和K-均值聚类算法的优势结构面分析

Dominant discontinuities based on stereographic projection and K-means clustering algorithm

  • 摘要: 对于结构面多而复杂的岩质边坡,其优势结构面的选取与分析是极其重要的工作。传统的结构面组数划分方法比较粗糙,分析结果有很大的主观因素,无法准确地给出结构面的优势产状,使得其结果在实际工程中使用不便。以结构面交线的筛选和分析为突入点,借助于赤平极射投影法,在楔形体滑移分析中首先确定可能的滑移区域,筛选出可能滑移的结构面交线,缩小计算范围,采用K-均值聚类算法和有效性检验,根据赤平极射投影分析得到滑移区域的对称轴中心作为初始凝聚点,通过多次迭代计算得到滑移区域内的优势结构面交线。将该方法用于重庆万盛黑山谷的岩质滑坡中,结果表明,将赤平极射投影与K-均值聚类算法相结合,计算得到的优势结构面交线分类合理,结果可靠,可以准确地确定结构面交线的优势产状。

     

    Abstract: For the rock slope with many complicated discontinuities, the selection and analysis of the discontinuities are a very important work. The traditional analysis methods are insufficient and inadequate. The analysis results have a lot of subjective factors, and are inconvenient to be applied in practice. The selection and analysis of intersection lines of discontinuity are treated as the starting point. With the aid of stereographic projection, the possible sliding zone of sliding wedge can be first determined. The possible sliding line of intersection can be selected. So the range of calculation can be reduced. The symmetric axis center is used as the initial rallying point. The dominant orientations of discontinuities can be selected and analyzed through the probabilistic analysis of the stereographic projection based on the K-means clustering algorithm. A case study on a rock slope in Chongqing is used. The study shows that the results are reliable and reasonable, and the dominant orientations and classification are more precise.

     

/

返回文章
返回