一种地球化学定量评价方法的改进

IMPROVEMENT OF A GEOCHEMICAL QUANTITATIVE ASSESSMENT METHOD

面金属量法是一种基于地球化学分析数据的资源量定量估算方法,该方法的应用主要局限于在数据层面上,结果只能反映数据异常本身的特点,而不是控制异常性质的地质条件。为解决这一局限性,文章引入相似系数的概念,相似系数的计算原理是将控矿地质因素数值化,从而转换成数据模型,通过某种数学算法与已知的模型区相类比,得出一个表示与模型区相似程度的数值,再用此相似系数对面金属量法估算的资源量进行校正,从而间接地实现控矿地质条件对定量预测结果的控制。该方法实现了纯数据驱动与智能驱动的有效结合,使得该定量方法更具科学性,同时增加了定量预测结果的可信度。从试验结果看,改进后方法的定量预测结果更加合理,具有实用价值。

Areal productivity is a quantitative mineral resources evaluation method based on geochemical analysis data. Its application is mainly confined to data calculation thus reflects characteristics of the data anomaly rather than the nature of geological condition that control the anomaly. This paper introduces the conception of similarity coefficient to solve the problem. The principle of coefficient calculation is to digi- talize the ore-controlling factors then transferred to data model. The model is correlated to a known model area by means of a mathematic algorithm to get the value that shows similarity with the model area. The similarity coefficient is used to calibrate the estimated results realizing indirect control of geological condi- tion on the quantitative prediction. This improved method effectively combined pure data-driven prediction with intelligence-driven prediction. Hence, the results become more reasonable and creditable, and the method worth to use widely in mineral quantitative assessment.