基于证据权法的成矿远景预测空间分析模型研究:以MAPGIS空间分析建模为例  

Spatial Analysis Model of Mineralization Prospect Forecast Based on Evidence Power Method:Take the MAPGIS Spatial Analysis Modeling as an Example

作　　者：陈文成[1] 

机构地区：[1]泉州师范学院资源与环境学院,福建泉州362000

出　　处：《地质科技情报》2011年第2期109-116,共8页Geological Science and Technology Information

基　　金：福建省自然科学基金项目(2008J0246);福建省教育厅A类科技项目(JA06017)

摘　　要：在证据权法成矿远景定量预测模型及其优劣势分析的基础上,提出了以多维空间线性变换方法改进证据权法,采用"舍弃法"解决证据独立性问题,进行了基于因子分析的公因子方差赋权与方差贡献赋权的模型修正,从而使正—负证据权赋值与"1-0"赋值等价,将证据权法拓展为变量构置、筛选及赋值(变换)的准则(方法)。将综合信息成矿远景预测概括为找矿信息提取与变量构置、筛选、赋值(变换)、赋权、多变量综合及其临界值确定、成果视觉化等环节,采用混合型空间模型法建模,使面向证据权法及其拓展的成矿远景定量预测MAPGIS空间分析模型更具普适性。通过1∶5万长坑幅铁、铜、铅锌综合信息成矿远景定量预测评价,实证了改进方法具更好的兼容性与应用效果,并圈定了铁、铜、铅锌成矿远景区13处(A、B、C类各2,4,7处),为该区进一步的地质找矿工作提供了依据。Based on the evidence power method prospect quota forecast model and the analysis of its positive and negative points,this paper proposes the multi-dimensional space linear substitution method to improve the evidence power and to solve the evidence independence problem by using ＂the rejection method＂.With the factor analysis on common factor variance bestows on the weight and the variance contribution bestows on power model revision,the positive-negative evidence power evaluation is equal with 1-0 evaluation so as to develop the evidence power method for the variable construction,screening and evaluation and transformation criterion/method.The integrated information mineralization prospect forecast can be summarized into such steps as mineralization information extraction,variable construction,sorting,evaluation/transformation,weighting,multivariable synthesis and marginal value determination and achievement visualization.The use of mixed space modeling to establish a model universalizes the evidence-power-oriented method and its developed mineralization prospect quota forecast on the MAPGIS space analysis model.The application of this method in the prediction of a 1∶50,000 long pit of iron,copper,lead and zinc confirms that the improved model has better compatibility and application effect,for 13 iron,copper,lead and zinc zones（2,4,and 7 zones respectively for class A,B,and C） are successfully identified.This has provided evidence for the mineral prospecting in the area.

关 键 词：成矿远景 定量预测 证据权法 空间分析模型 MAPGIS 

分 类 号：P612[天文地球—矿床学]