Research on a priori and posterior weighting methods for Multi-GNSS combined single point positioning
-
摘要: 针对多模全球卫星导航系统(GNSS)融合伪距单点定位随机模型难以精确构建的问题,在全球范围内选取了10个多GNSS实验跟踪网MGEX(Multi-GNSS Experiment)观测站连续7天的观测数据,将四大GNSS系统的观测值分为五类,比较了高度角模型、用户等效测距误差(UERE)模型及验后Helmert方差分量估计模型的伪距单点定位精度. 结果表明:在四系统融合伪距单点定位时,Helmert方差分量估计模型能提高定位精度,高度角模型定位精度优于UERE模型,其中基于高度角的Helmert方差分量估计模型结果最优.
-
关键词:
- 多系统组合 /
- 随机模型 /
- 伪距单点定位 /
- Helmert方差分量估计 /
- 先验定权
Abstract: When using multi-system code observations to conduct the pseudo range single point positioning, a reasonable stochastic model needs to be determined. In this paper, the datasets from ten multi-system stations MGEX(Multi-GNSS Experiment) on seven consecutive days are selected to compare the positioning performance of pseudo range single point positioning with the elevation-dependent model and the user equivalent range error (UERE) model, as well as the posterior Helmert variance component estimation model based on the two a priori models. The observations from the four Global Navigation Satellite Systems (GNSS) are divided into five groups. The results show that the positioning accuracy can be improved when adopting the Helmert variance component estimation model. The positioning accuracy of the elevation-dependent model is better than that of the UERE model. The Helmert variance component estimation model based on the elevation-dependent weighting strategy achieves the best positioning performance. -
表 1 各站7天平均可见卫星数及PDOP值
测站名 PDOP 平均可见卫星数 GPS GLONASS Galileo BDS-2 BDS-3 总数 AREG 1.0 9.1 5.9 6.7 0.9 5.4 28.1 CEDU 1.0 8.6 5.3 4.1 9.8 5.1 33.0 DJIG 0.9 9.7 5.9 6.9 7.3 5.7 35.5 HARB 1.0 8.9 5.9 6.5 5.0 5.2 31.5 KAT1 0.9 9.3 5.7 6.2 11.2 5.5 37.9 METG 0.9 9.5 7.1 7.0 4.4 3.8 31.7 MIZU 1.0 8.7 5.5 6.3 6.6 1.7 28.8 POL2 0.9 8.6 6.2 6.3 7.6 7.2 35.9 UCAL 1.0 8.7 6.7 6.5 1.6 3.5 27.0 UNB3 1.0 8.7 6.4 6.4 0.9 6.9 29.2 表 2 不同定权模型平均定位精度对比
方向 ele/m ele+hel/m 提升
百分比/%uere/m uere+hel/m 提升
百分比/%E 0.447 0.409 8 0.464 0.417 10 N 0.725 0.658 9 0.740 0.661 11 U 2.511 2.389 5 2.527 2.398 5 3D 2.672 2.529 5 2.695 2.540 6 表 3 UERE模型观测值方差具体构成
卫星 时刻 高度角 UERE/m2 $\sigma _{{\rm{ele}}}^2$/m2 $\sigma _{{\rm{URA}}}^2$/m2 $\sigma _{{\rm{ion}}}^2$/m2 $\sigma _{{\rm{trop}}}^2$/m2 G15 15:15:00 40°50′20″ 7.514 0.228 5.760 1.368 0.158 17:25:00 25°11′4″ 8.523 0.301 5.760 2.135 0.326 18:11:00 15°1′14″ 10.198 0.437 5.760 3.303 0.698 G20 02:25:30 11°54′41″ 39.209 0.526 5.760 31.965 0.959 21:37:00 50°8′58″ 8.835 0.207 5.760 2.749 0.120 23:03:00 80°36′17″ 8.813 0.181 5.760 2.796 0.076 E11 04:08:30 80°21′15″ 17.350 0.181 11.560 5.532 0.076 05:36:30 60°22′48″ 18.897 0.194 11.560 7.048 0.096 07:21:30 20°5′45″ 37.750 0.352 11.560 25.380 0.457 C08 01:58:30 40°34′42″ 15.591 0.228 5.760 9.443 0.160 07:02:30 70°55′4″ 11.758 0.185 5.760 5.730 0.082 08:17:00 50°9′23″ 13.690 0.207 5.760 7.603 0.120 -
[1] 高晓, 戴吾蛟. 基于方差分量估计确定GPS/BD2组合定位先验权比[J]. 大地测量与地球动力学, 2013, 33(2): 136-138. [2] 何帆, 高成发, 潘树国, 等. 方差分量估计在多卫星导航系统联合定位中的应用[J]. 导航定位学报, 2013, 1(3): 56-61. [3] 段举举, 沈云中. 基于方差分量估计的GPS/GLONASS组合单点定位[J]. 测绘通报, 2011(4): 4-6. [4] 何俊, 袁小玲, 曾琪, 等. GPS/BDS/GLONASS组合单点定位研究[J]. 测绘科学, 2014, 39(8): 124-128, 170. [5] 李文凯, 高俊强, 栗广才. 一种多系统组合单点定位随机模型的确定方法[J]. 测绘工程, 2017, 26(7): 60-65. [6] 张乾坤, 刘小生. 一种北斗混合系统单点定位随机模型的优化方法[J]. 大地测量与地球动力学, 2020, 40(7): 756-760. [7] PAN L, CAI C, SANTERRE R, et al. Performance evaluation of single-frequency point positioning with GPS, GLONASS, BeiDou and Galileo[J]. Survey review, 2017, 49(354): 197-205. DOI: 10.1080/00396265.2016.1151628 [8] 李征航, 黄劲松. GPS测量与数据处理[M]. 第3版. 武汉: 武汉大学出版社, 2016. [9] 戴吾蛟, 丁晓利, 朱建军. 基于观测值质量指标的GPS观测量随机模型分析[J]. 武汉大学学报(信息科学版), 2008, 33(7): 718-722. [10] 崔希璋, 於宗俦, 陶本藻, 等. 广义测量平差[M]. 第2版. 武汉: 武汉大学出版社, 2009.