基于变异函数的距离加权反比法  

Inverse Distance Weighting Method Based on Variogram

作　　者：牛文杰[1] 

机构地区：[1]中国石油大学(华东)机电工程学院,山东东营257061

出　　处：《东华大学学报（自然科学版）》2011年第3期362-367,共6页Journal of Donghua University(Natural Science)

基　　金：教育部科学技术研究重点课题资助项目(108084)

摘　　要：通过插值建立研究变量的数学模型是数据场可视化中至关重要的一步.根据距离加权反比法原理并结合变异函数的优良特性,提出了基于变异函数的距离加权反比法,讨论了其插值特性,给出了近似最大似然法估计协方差参数的原理.通过应用实例比较可知,基于变异函数的距离加权反比法具有效率高、能抵御数据中的噪声、处理数据丛聚效应能力强的特性,其插值效果与克里金法基本相同.With the interpolation of data visualization,establishing the mathematical model of the studied variables is a crucial step.According to the principle of the inverse distance weighting method and the excellent features of variogram,inverse distance weighting method based on variogram is proposed.Its interpolation characteristics are specifically addressed and the approximate maximum likelihood estimation principle of covariance parameters is given.Comparison between this method and others in application shows that the inverse distance weighting method based on variogram is efficient and can withstand the noise in the data,simultaneously it is good at clustering effect while processing data.The interpolation result is mostly as same as that of Kriging.

关 键 词：数据场可视化 变异函数 距离加权反比法 克里金法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]