矿体块段模型品位估值快速搜索算法研究

为提高矿体块段模型品位估值时邻近组合样数据搜索的计算效率,加快搜索速度,通过构建包含整个组合样数据空间范围的三维网格,利用三维网格将组合样划分至网格单元块中,计算总索引值和单元块内组合样累计数;然后确定覆盖搜索椭球体的移动网格,并计算移动网格单元块相对于包含椭球体中心的单元块的偏移量。在对块段模型单元块进行品位估值时,依据移动网格的偏移量数组、总索引值和组合样累计数,快速确定估值单元块的邻近组合样,进而通过八分象限、邻近数据个数和每个钻孔数据个数等约束条件限制,得到最终用于克里格或距离幂次反比法估值的组合样数据。该方法通过数据分区和预处理提高了邻近组合样的搜索速度,适用于精确模拟矿体边界的大数据量块段模型品位估值。

A Quick Search Algorithm for Estimating Grade of Rock Mass Block Model

To improve the computational efficiency and accelerate search of nearby composite data for grade estimation of block mode, a three?dimensional grid covering the spatial extent of composite data was constructed and used to divide the composites data into cells for calculating the total index value and cumulative number of data for unit cell. The moving grid covering the search ellipsoid was established, and the offset values of cells within the moving grid relative to the cell containing the center of ellipsoid were calculated. When estimating the grade values of block model, the offset value array of moving grid, total index values and cumulative number of composite data were utilized to obtain the nearby data of each estimating cell. Furthermore, the nearby data were confined by the filtration of octant, number of nearby data and number of data belonging to the same drill, resulted in the nearby composite data for estimation methods such as Kriging or inverse distance weight. This algorithm has increased the searching speed of nearby data by data partition and pre?processing process, and can be applied to the grade estimation for block model containing large quantity data by simulating the shape of orebody precisely.