Comparative Study on Spatial Interpolation Methods for 3D Modeling of Coal and Rock Layers Based on Borehole Data  

作　　者：Xiang GE Liangliang YANG 

机构地区：[1]Huainan City Construction and Development Planning and Design Institute Co.,Ltd.,Huainan 232001,China

出　　处：《Meteorological and Environmental Research》2022年第6期68-72,77,共6页气象与环境研究（英文版）

摘　　要：The patial interpolation of borehole data is an important means of stratigraphic structure to construct a three-dimensional model of coal strata,and the reasonable selection of an effective spatial interpolation method will directly affect the accuracy of three-dimensional modeling of the strata.To select an effective spatial interpolation method and improve the accuracy of 3D modeling of formations,four interpolation methods(the inverse distance weight interpolation algorithm,the local polynomial interpolation algorithm,the radial basis neural network interpolation algorithm and the kriging interpolation algorithm)were compared and analyzed.In particular,the methods of interpolation algorithm,interpolation surface,sample test error,and cross-validation error were used.The experiment of 13-1 seam coal in the Huainan mining area showed the spatial surface interpolation effect of the radial basis neural network interpolation algorithm(RBF)compared with the inverse distance weight interpolation algorithm(IDW),local polynomial interpolation algorithm(LPI)and kriging algorithm.The three interpolation methods have higher accuracy and are more suitable for surface interpolation of coal seams,which is of great significance for improving the accuracy of subsequent 3D modeling of coal seams.

关 键 词：DRILLING Spatial interpolation Coal-rock surface Cross-validation 

分 类 号：TD163.1[矿业工程—矿山地质测量]