基于卫星高度角模型的GNSS双差观测量方差估计  

Variance estimation of GNSS double differenced measurements based on the satellite elevation stochastic model

作　　者：杨汀[1] 杜久升[2] 陈宜金[1] 

机构地区：[1]中国矿业大学(北京)地球科学与测绘工程学院,北京100083 [2]河南理工大学测绘与国土信息工程学院,河南焦作454000

出　　处：《河南理工大学学报（自然科学版）》2017年第2期63-67,共5页Journal of Henan Polytechnic University（Natural Science）

基　　金：国家自然科学基金资助项目(41502227);河南省高等学校重点科研项目(16A42002)

摘　　要：GNSS定位精度及可靠性与数据处理中的随机模型密切相关,为了获得与实际相符的卫星高度角指数模型,使用实测数据,通过最小二乘方差分量估计方法对模型中的3个未知参数进行估计。算例结果分析表明,该随机模型的定位精度优于等权模型与常用的指数模型,同时指出若随机模型选取不当,定位精度将劣于等权模型。因此,在GNSS高精度数据处理中,应根据接收机性能和观测量类型,合理建立随机模型。In global navigation satellite system（GNSS） data processing,the correct stochastic modelcan be applied to improve the positioning accuracy and efficiency.In practice,the variances of raw GNSS measurements are associated with satellite elevation angles,least-squares variance component estimation is applied to evaluate three unknown parameters of the exponential model on the basis of double differenced GPS measurements collected by Trimble R8 receivers,and the positioning accuracy is analyzed compared with the identical weight model and the commonly exponential model.The results show that the proposed stochastic model is more accurate and the elevation model is not necessarily better than the identical weight model if the parameters are not appropriately specified.Therefore,it should choose the realistic stochastic model considering the type of receiver and observation.

关 键 词：GNSS 随机模型 高度角模型 指数模型 最小二乘方差分量估计 

分 类 号：P228.4[天文地球—大地测量学与测量工程]