多元统计方法在地气异常推断解释中的应用  

Application of Multivariate Statistical Methods in the Geogas Anomalies Inferred Interpretation

作　　者：刘国安[1] 周四春[1] 赵辉[1] 孙森[1] 刘俊[1] 

机构地区：[1]成都理工大学核技术与自动化工程学院

出　　处：《金属矿山》2013年第9期113-116,共4页Metal Mine

基　　金：国家自然科学基金项目(编号:41074094);国家"深部探测技术与实验"专项(编号:SinoProbe-03-01-4C);四川省教育厅重点科技项目(编号:10ZA118)

摘　　要：以湖南黄沙坪铅锌矿和云南施甸东山铅锌矿整装勘查区为对象,开展了地气测量探测隐伏多金属矿的实验研究。利用因子分析、聚类分析和判别分析方法对地气异常进行推断解释,获得了矿异常与矿体和下伏岩浆岩之间的联系,揭示了矿异常与非矿异常的差别,并说明综合运用多元统计方法可以更加准确地推断未知的地气异常。With Huangshaping lead-zinc mine in Hunan province and Dongshan lead-zinc integrated exploration area in Shidian of Yuanna province as the object, experimental research of geogas measurement to detect concealed muhi-metal mine is carried out. The factor analysis, cluster analysis and discriminant analysis methods are adopted to infer the geogas a- nomalies ,obtaining the connection relationship among the ores anomalies, ore-bodies and underlying igneous rocks. The difference of ore anomalies and non-ore anomalies is revealed. It is concluded that the integrated use of the multivariate sta- tistical method can more accurately infer the unknown abnormal geogas.

关 键 词：地气异常 因子分析 聚类分析 判别分析 

分 类 号：O212.4[理学—概率论与数理统计]