GNSS网解星间单差整数模糊度解算研究

Ambiguity resolution plays the role as a performance multiplier in GNSS applications with carrier phase observation. Along with the of invention of Precise Point Positioning(PPP) technique and the concept of PPP base on RTK Networks, the mode for GNSS precise positioning begin to return to the mode of point positioning with a single receiver, as it is original designed, correspondingly, the object of ambiguity resolution has been changed for the double-differenced ambiguities to the single/zero-differenced ones. Network solutions are the fundamental data processing mode which support the operation and various applications of GNSS. However, presently the products (especially the satellite clocks) solved with double-differenced ambiguity resolution provided by the overwhelming majority of Analysis Centers participating in the international GNSS service (IGS) and the international GNSS Monitoring and Assessment System(iGMAS) do not allow the user to perform single/zero-differenced ambiguity resolution. For this reason, in the current proposal, an approach of single-differenced ambiguity resolution in Network solutions is presented, and the research would be systematically carried out on three levels: basic theory, practical algorithm and applications & experiments with emphasis on mathematical proven and definition of the single-differenced integer ambiguity solution and the key techniques e.g. sequential fixing and datum selection for the single-differenced ambiguities. It is foreseen, this approach would provide solution of phase clock which allow single/zero-differenced ambiguity resolution for PPP users, and is superior to zero- and double-differenced ambiguity fixing respectively in algorithm stability and computation efficiency..The achievement of this proposal will not only develop and enrich the theory and method for GNSS data processing, but will also help the iGMAS to provide product and service distinguishing from the IGS.

整数模糊度解算是GNSS载波相位应用的效能倍增器，随着PPP技术及PPP－RTK概念的出现，GNSS精密定位开始回归到其最初设计的模式——单点定位，模糊度解算的研究对象也从“双差”向“单/非差”转移。网解是GNSS运行和广泛应用的基本数据处理模式，但是当前绝对大多数IGS和iGMAS分析中心基于双差模糊度固定的网解获得的卫星钟差不支持用户进行非/单差模糊度固定。本项目创新性地提出在网解中进行星间单差模糊度固定的方案，拟从“基础理论”，“实用算法”和“实验与应用”三个层面系统地展开研究。重点对单差整数模糊度解进行数学证明和定义，攻克单差模糊度的基准选择和序贯固定等关键技术。该方案可以解算出支持单/非差模糊度固定的相位钟差，其稳定性和计算效率要分别优于非差和双差模糊度固定。.本项研究不仅可以丰富和发展GNSS数据处理理论和方法，也可以助力iGMAS提供差异于IGS的产品和服务。

精密单点定位（PPP）的模糊度固定是近年来GNSS精密定位领域的研究热点。通过处理全球GNSS网的观测数据获得的轨道和钟差产品是进行PPP应用的基础。然而，当前绝对大多数IGS和iGMAS分析中心基于双差模糊度固定技术获得的卫星钟差不支持PPP的模糊度固定（需要额外的产品）。本项目提出并实现在GNSS网解中进行星间单差模糊度固定的理论和技术。从“基础理论”，“实用算法”和“实验与应用”三个层面系统地展开研究。在理论方面，提出更广义的关于模糊度参数整数性质的定义：一个参数可以表示为若干个整数的线性组合即称之为具有整数性质。将顾及模糊度整数性质的GNSS网数据处理问题抽象为广义的混合整数最优化问题。基于秩亏网平差理论，证明了星间单差整数模糊度的存在性及其条件。基于带约束条件的平差模型论述了非差、星间单差和双差整数模糊度解算方法的理论等价性。在算法方面，项目实现并验证了在网解中进行星间单差模糊度固定的算法和软件。利用图论中的最小生成树概念解决了星间单差模糊度固定的关键技术——独立星间单差整数模糊度基准的选择问题；提出模糊度连续的概念解决了卫星或测站观测中断条件下的基准模糊度选择问题。在实验和应用方面，课题组采用实测的GPS、BDS、Galileo观测数据验证了算法的正确性。其中，GPS实测数据实验表明，星间单差模糊度固定获得的测站坐标、卫星轨道和钟差的精度与传统双差模糊度固定解几乎一致甚至更优，且模糊度固定的效率更高，得到的卫星钟差支持PPP模糊度固定。项目还提出了基于星间单差模糊度固定技术的PPP算法和相位距离生成方法。项目的研究成果丰富了GNSS网解整数模糊度解算理论。该算法的应用使得课题组所开发的SPODS软件成为目前国际上唯一一个可以进行星间单差模糊度固定的GNSS高精度数据处理软件。