基于高斯模型的样本缺失GPS时间序列重构方法

GPS Time Series Reconstruction Method Based on Gaussian Model

缺失样本的存在会造成GPS时间序列速度估计的不确定性,从而影响GPS时间序列的应用。针对该问题本文提出一种基于高斯模型的样本缺失GPS时间序列重构方法,首先利用高斯概率密度函数对GPS时间序列的先验分布进行建模,在此基础上构建全概率贝叶斯统计模型,采用期望最大(Expectation Maximization,EM)算法对模型参数(隐变量)进行迭代更新并计算其最大似然估计值,最终完成信号重构。分别对随机缺失和分段连续缺失两种情况进行实验分析,结果表明所提方法相对于传统插值方法可以获得更好的重构性能。

The existence of missing samples will cause the uncertainty of GPS time series speed estimation,thus affecting the application of GPS time series.In this paper,a method of GPS time series reconstruction based on Gaussian model is proposed.Firstly,Gaussian probability density function is used to model the prior distribution of GPS time series.On this basis,a full-probability Bayesian statistical model is constructed,and the model is solved using the Maximum Expectation(EM)algorithm,and the random missing and segmented fixed missing are experimentally analyzed.The results show that method proposed in this paper can obtain better reconstruction performance than the traditional interpolation method.