基于多源数据融合的卫星导航空间信号完好性处理方法研究

The performance of GNSS consists of accuracy, integrity, continuity and availability. Integrity means the ability of alerting when an unplanned fault occuring and the positoning error exceeding the alert limit. As accuracy of positioning is related with the accuracy of signal-in-space, the integrity of positioning is related with the integrity of signal-in-space too. The faults affecting on integrity may come from some factors such as navigation satellite orbit, clock, and signals, etc. Each fault factor has different characteristic. The fault may be caused by one single factor or more than one. Therefore, in order to ensure the integrity of signal-in-space, some kinds of monitoring become the focuses in developing satellite navigation system in recent years, which are ground-based monitoring, satellite autonomous monitoring and inter-satellite-link monitoring. The theory of fusion processing, model-based fault diagnosis and the protect level calculating method has been applied widely, it is obvious that fusion processing based on multi-sources may work much better than the monitoring method based on one single source, so that the integrity monitoring of sigal-in-space is based on multi-source is the development trend. The research is foucused on the analysis of the characteristic of the faults, the error modelling method, biulding different kinds of detectors to detect different kinds faults, model-based fault diagnosis and threshold computation, multi-sources-fusion algorithm, so as to improve the ability of fault detection and performance of integrity, and ensure the creditability of the signal-in-space.

卫星导航系统的应用性能要求包括精度、完好性、连续性、可用性等。定位精度性能是由导航空间信号精度决定的，与此相对应，定位完好性性能是由导航空间信号完好性性能决定的。空间信号故障有卫星轨道、卫星钟、信号生成等各种因素，这些故障因素具有不同物理特性，且各种故障类型有可能单一或同时发生。如何利用系统地基监测、卫星自主监测、星间链路监测等模式，较好地实现系统完好性性能保证，是近年来卫星导航系统发展与研究的热点问题。若将多源数据融合处理方法、基于模型的故障诊断与阈值确定理论引入多模监测数据综合处理，势必会取得整体最优效果。为此，有必要开展卫星导航系统空间信号故障特征分析和模型化研究，构建适应不同故障特征的异常检测器，在局部观测空间对特定故障形成最优或次优估计的同时，按某种最优融合准则形成系统整体状态的最优估计，以提高空间信号完好性故障状态的检测识别性能，确保卫星导航系统精确可信。

课题深入、系统地研究了卫星导航空间信号完好性处理理论与方法，利用长期观测数据对卫星空间信号故障特征进行了分析和建模，并将相应研究成果应用到北斗新一代试验卫星轨道与钟差确定中，取得了良好的应用效果；提出了系统级多源完好性监测技术体系，详细设计了完好性监测处理流程、方法与参数化表达，形成了技术体制设计；将抗差估计等理论引入到RAIM算法中，实现了多星座多源数据条件下的多故障检测与排除；基于完好性概率分配构建了空间信号完好性与用户定位完好性的状态转换方法；提出了多系统条件下的ARAIM算法和基于验后方差的用户保护水平（PL）算法，实现了基于有限样本的完好性评价方法。.课题研究成果主要包括：系统建立了基于多源数据融合的卫星导航空间信号完好性处理理论与方法，深入分析研究了卫星导航空间信号故障特性及其模型化，构建了系统级的完好性监测技术体系，提出了多星座多源数据条件下的用户端自主完好性监测算法和有限样本下的完好性评价方法。