核主成分分析在三维综合信息成矿预测中的应用研究

hree-dimensional comprehensive information is one of the key technique to three-dimensional metallogenic prediction. In this paper, the author taked the Tonglushan Copper Iron Mine in Hubei Province as the study area. Based on studying on the data of actual exploration and production and summaring of metallogenic regularity, the author carried out the quantitative study on the exacted of the 11 ore-controlling from geological and geochemical prospecting information. At the same time, three-dimensional geological model was build. On this basis, 3D geological prospecting model was build combined with kernel principal component analysis. Mineralization favorable degrees of prediction 1, 2, 3, 4, 11 and 14 unit were highest through analyzing the 15 prediction units divided. The predicting results were highly tallies with the actual data contrasted with the proepecting results. The application results showed that it was effectively realized to locate the concealed ore bodies by the 3D metallogenic prediction model established in this paper.