安卓智能手机的精密单点定位模型研究

Research on precise point positioning model based on Android smartphone

针对当前智能手机的全球卫星导航系统(GNSS)原始数据质量不佳导致常规精密单点定位(PPP)模型无法有效利用以及缺少智能手机PPP模型测试分析资料等问题,该文以载波加钟差的改进精密单点定位模型为基础,扩展至相应的载波加钟速模型、UofC模型、无模糊度模型、载波二次差模型及星间差分模型,简述了各个模型的特点,并利用华为P10智能手机的GNSS原始数据,测试并分析了各个模型定位精度情况:星间差分PPP模型在短期内(10 min)定位精度最优,平面中误差在1.8 m左右,高程2.1 m左右;改进模型与载波加钟速模型定位精度相当,UofC模型精度稍差;无模糊度模型与载波二次差模型需在Kalman滤波过程中采用降权方式防止其发散,定位精度取决于前期历元的数据质量。

In view of the current smartphone global navigation satellite system(GNSS)inferior quality of the raw data lead to conventional precise point positioning(PPP)model can not effectively use and lack of test and analysis data of smartphone PPP model,based on the improved PPP model of carrier plus clock error,this paper extended to the corresponding carrier-clock-speed model,UofC model,unambiguity model,carrier-quadratic-differenced model and satellite-differenced model,and briefly describes the characteristics of each model.Using the GNSS raw data of Huawei P10 smartphone,the positioning accuracy of each model were tested and analyzed:the satellite-differenced PPP model had the best positioning accuracy in the short term(10 min),the RMSE in the plane was about 1.8 m,and in elevation was about 2.1 m.The improved model was comparable to the carrier-clock-speed model,and the UofC model was slightly less accurate.The unambiguity model and the carrier-quadratic model needed to use the decentralization method in the Kalman filtering process to prevent divergence.Positioning accuracy depended on the data quality of the previous epoch.