矿产资源估算中定量邻域分析与优化研究  

Quantified kriging neighborhood analysis and optimization in mineral resource estimation

作　　者：刘金辉[1,2] 王京彬[1] 李峰[2] 张汉成[3] 王翔[4] 

机构地区：[1]北京矿产地质研究院,北京100012 [2]昆明理工大学,云南昆明650500 [3]有色金属矿产地质调查中心,北京100012 [4]中国地质调查局发展研究中心,北京100037

出　　处：《中国矿业》2016年第3期154-157,共4页China Mining Magazine

摘　　要：在应用线性、非线性地质统计学进行矿产资源/储量估算或品位条件模拟过程时,邻域分析是其中非常关键和重要的一个环节。通常情况下,在进行资源/储量估算中邻域参数通过人为控制和设定,这导致在一定程度上具有主观因素。研究认为估计邻域的大小,应由各个方向变异函数变程及搜索椭球体等参数来确定,优化邻域需选择足够多的样本,且得到最大回归斜率,同时具有最小负克里金权重。克里金邻域参数的合理设置将直接影响最终块体估值结果,这为资源量估算和矿床评价提供了理论和实践基础。During the mineral Resources/Reserves estimation by linear,non-linear geostatistics as well as mineral grade conditional simulation,neighborhood analysis is of the very critical and important process.However,the neighborhood parameters for the mineral Resources/Reserves estimation are commonlyto be defined manually,which has subjective factorsin part and might lead to the estimated value mismatch with actual value,once the neighborhood parameters are to be set arbitrarily.The searching distance of neighborhood should be defined by the directional variogram and search ellipsoid,and the optimized neighborhood is to search sufficient samples,and get maximum regression of slop with the minimum negative kriging weight.The parameters of kriging neighborhood directly affect the final block estimated value,therefore,selecting reasonable and optimized neighborhood parameters will provide the theoretical and practical basis for the mineral resource estimation and mineral deposit valuation.

关 键 词：地质统计学 回归斜率 克里金方差 邻域优化 

分 类 号：TD15[矿业工程—矿山地质测量]