基于空间结构性的距离幂次反比法品位估值研究  

IDW Estimation for Ore Grade Based on Spatial Structure of Data

作　　者：荆永滨 孙光中 毕林[3] Jing Yongbin;Sun Guangzhong;Bi Lin(School of Safety Engineering, Henan Institute of Engineering, Zhengzhou 451191, China;State Key Laboratory of Safety and Health for Metal Mines, Ma＇anshan Anhui 243004,China;School of Resources and Safety Engineering, Central South University, Changsha 410083, China)

机构地区：[1]河南工程学院安全工程学院,郑州451191 [2]金属矿山安全与健康国家重点实验室,安徽马鞍山243004 [3]中南大学资源与安全工程学院,长沙410083

出　　处：《地质科技情报》2018年第2期215-220,共6页Geological Science and Technology Information

基　　金：国家自然科学基金项目(41572317);金属矿山安全与健康国家重点实验室开放基金项目(2017-JSKSSYS-02);河南省科技攻关项目(172102310643);河南省科技开放合作项目(172106000073)

摘　　要：为使距离反比加权法(IDW)的估值结果反映出矿石品位的空间结构性,提出了基于数据空间结构性的距离反比加权估值方法(SIDW)。对钻孔样品按长度进行品位组合,通过计算矿体走向、倾向和厚度方向的变异函数分析组合样品中矿石品位的空间结构性。根据变异函数的方向和变程建立表示结构性的各项异性椭球体模型,利用线性变换矩阵将样品的品位数据的各向异性转换为各向同性,从而将样品搜索范围从与矿体产状一致的椭球体转换为球体,进而将计算权重的距离转换为各项同性变换后的距离。以某铜矿为例,分别将IDW和SIDW的估值结果与普通克里格法的估值结果进行对比分析,3种方法的品位估值结果的平均值和标准差基本一致,但SIDW与普通克里格法估值结果的相关系数、相对误差的平均值和标准差等指标均优于IDW,说明基于数据空间结构性的距离反比加权法与普通克里格法的品位估值结果更接近,能够以简单的方法较大程度地反映矿石品位的空间变化,得到较好的矿石品位估值结果。The structured inverse distance weighting （SIDW） method is proposed to make the estimate re- sults of inverse distance weighting （IDW） method reflect the spatial structure characteristics of ore grade. Borehole samples were composited by length, and then the spatial structure of ore grade was analyzed by calculating the variograms on orebody strike, direction of dip and thickness. Model of anisotropic ellipsoid for spatial structure was established according to the orientations and ranges of variograms, and next, the feature of sample data was transformed from anisotropism to isotropism by using the linear transformation matrix. Thereby, the search range of IDW method was transformed from ellipsoids with the same orienta- tion as the ore deposit to spheres. Moreover, the distances for weights were transformed into isotropic dis- tances. A copper mine was taken as an example. IDW and SIDW were compared with the ordinary kriging in estimate results, respectively. Results show that the mean value and standard deviation of estimate re- sults are almost the same for the three methods. But the correlation coefficient, mean value of relative er- ror and standard deviation of both SIDW and ordinary kriging are superior to IDW. The comparison shows that SIDW method is much closer to the ordinary kriging method in the estimation results. The SIDW method can largely reflect the spatial variation of ore grade and achieve more accurate grade estimation re- sults in a simple way.

关 键 词：品位估值 距离反比加权 变异函数 各向异性 

分 类 号：P624.7[天文地球—地质矿产勘探]