作为地质数据处理过程中经常用到的一种空间插值方法,距离幂次反比(IDW)法计算过程中如何设置有效参估样品是这种方法一直存在的问题。通过枚举的思路,基于交叉验证方法对所有可能的参估样品个数取值进行遍历和判断,自动优选出该方法框架下的最优参估样品数据集。利用这种方法对实际数据的插值过程进行了分析测试,所得计算结果与普通克里格法的精度较为接近,明显优于常规经验性参数设置方法的估计值,表明该方法具有有效性和实用性。
As a typical spatial interpolation method in geosciences data processing,inverse distance weighting(IDW)methods have a long-time standing problem which is how to choose the best participant samples while calculating.This paper proposes an alternative way based on enumeration method.By orderly analyzing and estimating the effectiveness of every possible participant dataset in cross-validation,this method can automatically optimize the proper samples.Additionally,the calculation process with a real dataset by the proposed method is illustrated and analyzed by a case study,from which the calculation result is clearly superior to conventional method.Thus,both theoretical analysis and practical application can show the validity and practicality of the proposed method.