一种融合多方向变异性的改进克里金插值算法

An Improved Ordinary Kriging Algorithm Fusion of Multidirectional Variability

在室内位置指纹定位技术中,采用克里金插值算法可有效降低指纹数据库构建的工作量。但是,现有算法通常只考虑了信号在一维方向上的变异性,没有考虑室内环境中信号在多个方向存在的变异性。为此,本文提出了一种融合了多个方向信号变异性的改进克里金插值算法,充分考虑了二维室内环境中信号在不同方向的变异性对插值精度的影响。实验结果表明,与克里金插值算法、距离反比加权插值算法等算法相比,本文算法对指纹数据的插值精度更高。

In the practice of indoor fingerprint localization technology,the application of Kriging interpolation algorithm can efficiently reduce the workload of fingerprint database establishment.However,the Kriging Interpolation algorithm usually only cover the signal variability in one dimension direction,the variabilities of indoor signals in different dimension direction are not considered.Therefore in this paper,an improved Kriging interpolation algorithm which fuses signal variability in different direction is proposed.The effects of signal variabilities in two dimension indoor environment to the accuracy of interpolation are fully considered.Experiment results show that compared with Kriging interpolation algorithm,Inverse-distance weighting algorithm,the proposed algorithm can get more precisely interpolation data of fingerprint.