多频GNSS非差非组合精密数据处理理论及其应用

发布时间：2018-05-21 19:19

本文选题：多频多模 + 统一模型　；参考：《武汉大学》2013年博士论文

【摘要】：随着我国BDS系统的运行服务、GPS与GLONASS系统的现代化,多导航系统(GPS、 GLONASS、Galileo、 BDS等)、多频率(两个以上)、多类型(P码、C码等)观测信号是GNSS系统发展的重要特征与优势。更为丰富的观测信号、更多类型的地面跟踪站网资源,需要相应的高精度数据处理理论方法作为支撑,才能充分发挥多频GNSS的优势。然而,现有以GPS双频观测为基础建立的高精度数据处理理论、算法模型以及软件系统已不能完全满足多系统多频高精度数据处理的需求,包括IGS组织在内的国内外研究机构都尚未提出完整的解决方案,多系统多频GNSS高精度数据处理已成为卫星导航研究领域的热点问题。另一方面,我国北斗系统是第一个提供服务的具有全星座三频信号调制的卫星导航系统,如何充分发挥三频观测的优势,是提升北斗系统竞争力的关键因素之一。对多系统多频GNSS高精度数据处理开展研究,不仅促进北斗系统的应用优势,同时将进一步发挥北斗卫星导航系统对GNSS高精度数据处理理论方法的贡献。本文围绕实现融合各类导航系统、涵盖不同频率信号、兼顾多样化应用需求的高精度数据处理的研究目标。提出了采用非差非组合观测量的GNSS多频高精度数据处理理论,建立非差非组合卫星钟差与时延偏差处理方法、非差非组合电离层延迟处理方法、非差非组合整数模糊度固定方法。研制了多频GNSS数据处理软件,实现GPS/BDS多系统多频高精度数据处理应用。通过对信号时延偏差、钟差、电离层等网解处理分析以及单站定位精度比较,验证了非差非组合处理算法模型的灵活性与正确性。本文具体工作和主要贡献包括： 1.从GNSS原始观测信号出发,提出定义信号特征性时延偏差,形成时间基准统一的非差非组合观测方程,建立了多频GNSS处理统一数学模型表达,论证通过参数转换与基准引入实现新模型的特定应用,从理论上分析了模型方法的普适性与灵活性。 2.建立了非差非组合卫星钟差与时延偏差处理方法,针对非差非组合模型中钟差与时延参数相关性,提出了满足最小约束解的模型正则化方法,实现了钟差与时延偏差去相关。分析给出了不同频率观测值数据处理中钟差与时延偏差产品的组合应用方法。比较了基于传统无几何距离组合的硬件延迟估计以及基于无电离层组合的卫星钟差估计模型中隐含的基准条件,论证了其与本文最小约束解基准条件的兼容性。 3.提出了非差非组合电离层延迟处理方法,该方法综合顾及电离层频间约束、时空变化约束以及电离层延迟先验改正信息,建立单站电离层参数化模型。基于空间统计学理论确定了单站电离层时空相关约束的水平梯度与随机模型,采用虚拟观测值形式引入先验电离层模型延迟改正量。从模型上论证了附加电离层先验信息的非差非组合模型是现有无电离层组合模型的扩展。结合精密单点定位算例分析,验证了新模型能显著提高单/双频精密单点定位精度,与无电离层组合算法相比,NEU方向精度提高分别为：单频47.8%、53.7%以及52.5%；双频：16.5%、13.9%以及16.8%；结合电离层建模算例分析,验证了中国陆态与省级CORS网电离层模型,与IGS全球电离层图比较,精度分别提高50.3%、70.7%。 4.通过对非差非组合模型中UPD和模糊度参数解空间分析,提出了一种基于迭代处理策略的基准引入方法,并采用该方法获得了非差非组合模糊度。在此基础上,指出需要将其转换为宽巷/窄巷模糊度以避免电离层残余误差对模糊度取整及UPD分离影响,推导了三频/双频条件下非差非组合模糊度(UPD)与超宽巷/宽巷/窄巷模糊度(UPD)的相互转换关系。通过覆盖中国区域的GPS UPD估计与非差非组合模糊度固定算例验证了本文算法的可靠性,其中相对于浮点解,固定解NEU方向定位精度分别提高了5.6%,8.6%以及11%。 5.实现采用非差非组合数据处理模型的GPS、BDS实测数据处理应用,通过基于网解处理模式的未校正偏差估计、高精度卫星钟差解算、区域电离层建模以及基于单站处理模式的单/双/三频浮点解、固定解算例分析,进一步比较验证了多频GNSS统一解算模型的优越性以及处理软件的可靠性。
[Abstract]:With the operation service of BDS system in China, the modernization of GPS and GLONASS system, multi navigation system (GPS, GLONASS, Galileo, BDS, etc.), multi frequency (more than two), multi type (P code, C code) observation signal is an important feature and advantage of the development of GNSS system. More abundant observation signals, more types of ground tracking station network resources, need corresponding The high precision data processing theory and method can give full play to the advantages of multi frequency GNSS. However, the existing high precision data processing theory, algorithm model and software system based on GPS dual frequency observation can not fully meet the needs of multi frequency and high-precision data processing, including the domestic and foreign research of IGS organization. Research institutes have not yet put forward a complete solution. Multi system and multi frequency GNSS high precision data processing has become a hot topic in the field of satellite navigation.
On the other hand, the Beidou System in China is the first satellite navigation system with full constellation tri frequency modulation. How to give full play to the advantages of the three frequency observation is one of the key factors to enhance the competitive power of the Beidou system. The research on the multi frequency and multi frequency GNSS high precision data processing not only promotes the application of the Beidou system. At the same time, it will further give play to the contribution of Beidou satellite navigation system to the theory and method of GNSS high precision data processing.
This paper focuses on the research goal of high precision data processing which combines various kinds of navigation systems, covering different frequency signals and taking into account the diverse application requirements. A GNSS multi frequency and high precision data processing theory is proposed, which uses non differential non combinatorial view measurement, and the non differential non combinatorial satellite clock difference and delay deviation processing method and non differential non combinatorial ionization are established. The layer delay processing method, the non differential non combinatorial integer fuzzy degree fixed method. The multi frequency GNSS data processing software is developed to realize the GPS/BDS multi frequency and multi frequency high precision data processing application. Through the analysis of the signal delay deviation, the clock difference, the ionosphere and other network solutions and the single station positioning precision comparison, the non difference non combinatorial processing algorithm model is verified. The specific work and main contributions of this article include:
1. from the original observation signal of GNSS, the characteristic time delay deviation of the signal is defined and the non difference non combination observation equation of the time datum is formed. The expression of the unified mathematical model of the multi frequency GNSS processing is set up, and the specific application of the new model is proved by the parameter conversion and the reference. The universality of the model method is theoretically analyzed. Flexibility.
2. the processing method of clock difference and delay deviation for non differential non combinatorial satellite is established. In view of the correlation between clock difference and time delay parameter in non differential non combinatorial model, a model regularization method is proposed to satisfy the minimum constraint solution. The correlation between clock difference and time delay deviation is realized. The analysis of clock difference and time delay deviation in different frequency data processing is given. The combined application method of the product is used. The hardware delay estimation based on the traditional non geometric distance combination and the reference condition of the satellite clock difference estimation model based on the non ionospheric combination are compared, and the compatibility with the minimum constraint solution base condition of this paper is demonstrated.
3. a non differential non combinatorial ionospheric delay processing method is proposed. This method takes into account the ionospheric frequency constraints, spatio-temporal change constraints and the ionospheric delay prior correction information, and establishes a single station ionospheric parameterized model. Based on the spatial statistics theory, the horizontal gradient and random model of the spatio-temporal correlation constraints of the single station are determined. The model of the transcendental ionosphere model is introduced by the virtual observational value form. The non difference non combinatorial model of the additional ionospheric prior information is proved to be the extension of the existing non ionospheric combination model. It is proved that the new model can improve the precision of single / double frequency precision single point positioning and no ionosphere with the precision single point location analysis. Compared with the combination algorithm, the NEU direction accuracy is improved respectively: single frequency 47.8%, 53.7% and 52.5%; dual frequency: 16.5%, 13.9% and 16.8%. Combined with ionospheric modeling, the model of China land state and provincial CORS network ionosphere is verified. Compared with IGS global ionospheric map, the accuracy is increased by 50.3%, 70.7%., respectively.
4. by analyzing the spatial analysis of UPD and fuzzy parameter solutions in the non differential non combinatorial model, a benchmark introduction method based on iterative processing strategy is proposed, and the non differential non combinatorial fuzzy degree is obtained by this method. On this basis, it is pointed out that it needs to be converted into a wide lane / narrow lane fuzziness to avoid the fuzziness of the ionosphere residual error. The relationship between the non differential non combinatorial fuzzy degree (UPD) with the ultra wide alley / wide lane / narrow lane ambiguity (UPD) under the three frequency / double frequency conditions is derived. The reliability of the algorithm is verified by a fixed calculation example of the GPS UPD estimation and the non difference non combinatorial fuzzy degree covering the Chinese region, which is compared with the floating point solution and the fixed solution of the NEU direction. The positioning accuracy is increased by 5.6%, 8.6%, and 11%., respectively.
5. the application of GPS, BDS measured data processing using non differential non combinatorial data processing model, uncorrected deviation estimation based on net solution processing mode, high precision satellite clock error calculation, regional ionospheric modeling and single / double / triple frequency floating point solution based on single station processing mode, fixed solution example analysis, and further comparison and verification of multi frequency GNSS The advantages of the unified solution model and the reliability of the software are also discussed.
【学位授予单位】：武汉大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：P228.4

【参考文献】