A detailed stochastic model based on the elevation angle of BDS-3 satellite

In BDS satellite navigation and positioning, a reasonable random model of observation information is an important foundation for achieving high-precision positioning. This article is based on the limitation of using empirical values in traditional height angle random models. The least squares variance component estimation method is used to fit the parameters of the height angle random model in real time. The simplified Helmert variance component estimation method is used to estimate the accuracy of BDS-3 satellite observation information. Finally, the positioning accuracy of the improved height angle random model is compared in the high-precision relative positioning algorithm. The experimental results show that the improved height angle random model has a certain improvement in positioning accuracy compared to traditional height angle models, with improvements of 20.2%, 34.2%, and 14.7% in E, N, and U directions, respectively.