空间矿化模式的识别与鉴别

使用现代信息技术，基于空间数据圈定矿化分布和找矿目标，是当前矿产勘查的一个新方向。采用模糊聚类方法，以地质、地球物理和地球化学图层为输入数据，在GIS平台上对空间矿化模式进行识别，并通过地质解读从中鉴别出矿化模式。模糊聚类是基于c-means聚类的人工神经网络方法，以模糊隶属度作为不确定性推理指标，是一种非监督分类器。用这一方法进行空间矿化模式识别，可以在多种矿床类型共存和交叉重叠分布的环境中，识别并鉴别出个性化的矿化模式。以内蒙古乌兰浩特1:20万图幅为例，展示了模糊聚类空间矿化模式的识别与鉴别过程，取得了良好的效果。

Recognition and identification of spatial mineralization patterns

A recent development in mineral exploration is the use of information technology to define the distribution of mineralization and prospecting targets using spatial data.With geological, geophysical, and geochemical map layers as input data, fuzzy clustering was used in this study to recognize spatial patterns on the GIS platform, and the mineralization patterns were identified by geological interpretation.Fuzzy clustering is a hybrid spatial pattern recognition method composed of c-means clustering, neural network and fuzzy sets.Being an unsupervised classifier, it uses fuzzy membership as an uncertain measurement of the spatial pattern recognition.In the complex distribution environment where diverse deposit types coexist and intersect, different individualized mineralization patterns can be recognized and identified.This study illustrates the technique of fuzzy clustering spatial mineralization pattern recognition and identification, with successful results, using the 1:200 000 map of Ulanhot, Inner Mongolia as an example.