GNSS非差非组合数据处理与PPP-RTK高精度定位

Abstract

本文首先回顾了GNSS差分和组合数据处理的起源、特点和应用, 并阐述了其在多频多模背景下的局限。然后, 引出了非差非组合数据处理的诸多优势, 介绍了构建满秩非差非组合函数模型的消秩亏方法。基于该方法, 本文系统构建了系列非差非组合PPP-RTK模型, 包括伪距加相位和仅用相位两大类。两类模型均考虑不同的大气约束而衍生出电离层加权、浮点和固定3种变体, 且所有模型均顾及码分多址和频分多址两类系统。最后, 本文测试分析了非差非组合PPP-RTK在无人船、无人机和农机应用中的动态定位性能。试验结果表明, 3个场景下的模糊度首次固定时间均在10 s以内, 模糊度固定成功率在96%以上, 水平定位精度优于2 cm, 高程定位精度优于5 cm。在Galileo+GPS+BDS三系统农机定位中, 仅用相位PPP-RTK与伪距加相位PPP-RTK定位性能相当。与Galileo+GPS双系统定位相比, 三系统PPP-RTK将模糊度首次固定时间从几百秒缩短至几秒, 模糊度固定成功率从85%左右提升至99%以上, 定位精度提升了30%左右。

Alternate abstract:

This work discusses the origins, features, and applications of GNSS differenced and combined data processing methods. After pointing out some limits of the differenced and combined methods under the background of multi-frequency and multi-GNSS, we summarize the advantages of the undifferenced and uncombined data processing method and introduce the rank-deficiency elimination strategy that is used to construct full-rank undifferenced and uncombined models. Based on this strategy, we systematically formulate a class of undifferenced and uncombined integer ambiguity resolution-enabled precise point positioning (PPP-RTK) functional models, including code-plus-phase and phase-only categories. Both categories impose different constraints on ionospheric delays and yield three variants, including ionosphere-weighted, ionosphere-float, and ionosphere-fixed variants. All variants consider both code and frequency division multiple access signals. Finally, we evaluate the undifferenced and uncombined PPP-RTK performance by conducting boat-borne, airborne, and tractor-borne positioning experiments. Results show that, for three cases, the time to first fix is less than 10 s, the ambiguity success rate is higher than 96%, the horizontal positioning accuracy is better than 2 cm, and the vertical positioning accuracy is better than 5 cm. For Galileo+GPS+BDS triple-system tractor-borne positioning, the performance of phase-only PPP-RTK is comparable to that of code-plus-phase PPP-RTK. Compared to Galileo+GPS dual-system positioning, triple-system positioning decreases the time to first fix from several hundreds of seconds to several seconds, improves the ambiguity success rate from approximately 85% to better than 99%, and improves the positioning accuracy by approximately 30%.