基于模糊 C均值聚类和图形技术的结构面产状分析方法

针对传统的利用极点等密度图和玫瑰图的结构面分组方法主观性强和聚类分析方法不够直观的缺点，建议利用模糊C均值( FCM)聚类的隶属度的结果，结合图形技术绘制隶属度等值线图来进行结构面分组。隶属度等值线图充分利用了模糊C均值聚类中隶属度的信息，展现每个聚类的隶属度的空间分布规律，并且可以分辨出因随机因素形成的结构面，还可以直观地读出聚类中心的范围。三山岛金矿的实例证明，该方法同时具有传统方法直观和聚类分析方法客观的优点，并且能够适应优势组不明显的数据。

FCM and graphics technique based method for discontinuities occurrence analysis

ABSTRACT To solve the disadvantages of strong subjectivity for traditional plot methods of grouping discontinuities, such as the pole isodensity map and the occurrence rose graph, and the lack of intuitionism for popular clustering methods, this article introduces a plot method called the membership contour map. Based on the data of the membership matrix obtained through the fuzzy C-mean ( FCM) algorithm, the membership contour map is realized by a graphics technique. Due to the full use of membership information in FCM clustering, the membership contour map can show the spatial distribution of the membership degree of each clustering, distin-guish discontinuities caused by trivial random factors, and read out clustering centers by the scope form from the membership contour map directly. An application of Sanshandao Gold Mine proves that the membership contour map holds the advantages of both intuition-ism and objectivity, and can adapt discontinuities data, which do not have obvious dominant groups.