基于多模GNSS参考站网的实时区域/广域对流层延迟建模方法研究

A regional or wide-area tropospheric delay model with high accuracy and stability plays an important role in improving the accuracy and efficiency of real-time GNSS positioning. With respect to regional tropospheric delay modeling, the customary modeling method has the problems of insufficient information utilization, and high dependence on the stability of real-time data streams. In regard to the wide-area tropospheric delay modeling, the existing modeling method (mainly for empirical model) cannot capture the short-term variation of the tropospheric delay. For these reasons, we derive two new modeling methods to further improve the accuracy and stability of the existing regional and wide-area tropospheric modeling, respectively. One is named “A regional tropospheric delay modeling method using robust Extended Kalman Filtering algorithm (EKFTROP)”, and the other is named “A revised wide-area Grid-based tropospheric delay modeling method (GTP-R)". The main feature of the EKFTROP is that the model parameters are estimated recursively epoch-by-epoch by using robust Extended Kalman Filtering algorithm, which can fully exploit the information in time-domain, and make the modeling process more stable. The GTP-R can model the tropospheric delays with high accuracy and efficiency in wide-area and complex terrain conditions using large-scale distributed GNSS networks. The outcomes of relevance will further enrich the theory and methods of tropospheric delay modeling, and provide technique support for application of BeiDou regional and wide-area augmentation systems.

高精度、强稳健性的实时区域和广域对流层模型对提高GNSS定位精度及效率具有重要的作用。针对区域对流层建模，现有建模方法存在时间域信息利用不充分、受实时数据流稳定性影响大的问题；针对广域对流层建模，现有建模方法（以经验对流层模型为主）存在无法精细化模拟对流层延迟短时变化的问题。为此，本项目拟提出两种能分别提升现有实时区域和广域对流层建模精度及稳定性的新方法：“基于抗差卡尔曼滤波算法的区域对流层建模方法EKFTROP”和“改进的广域对流层格网模型建模方法GTP-R”。EKFTROP的特色是采用抗差卡尔曼滤波算法实现区域对流层模型参数的实时逐历元递归估计，使得建模信息利用更加充分，过程更加稳健。GTP-R的特色是基于大规模、分布式GNSS参考站网，实现广域范围、复杂地形条件下对流层延迟的高效率与高精度模拟。相关成果将进一步拓展现有对流层建模的理论，并可为我国北斗增强系统的建设及应用提供技术支撑。

对流层延迟误差已成为制约北斗/GNSS应用精度及效率进一步提升的主要瓶颈之一。现有区域对流层建模方法存在时间域信息利用不充分、对实时GNSS产品和观测数据流稳定性强依赖的问题；现有广域对流层模型存在无法精细化模拟对流层短时变化和复杂地形应用精度低的问题。为此，针对区域建模，本项目提出一种基于抗差卡尔曼滤波的建模方法，充分利用了区域对流层的短时平稳变化特性，并引入抗差估计策略，显著提升了区域对流层建模的稳健性。针对广域建模，本项目建立了一种基于数值天气预报模式的实时广域对流层格网模型，并根据对流层干、湿分量随高度的变化特性，分别研制了新的高差归算方法，显著提升了格网模型在复杂地形，特别是大高差场景下的应用精度。本项目成果进一步丰富了对流层建模的理论与方法，为我国北斗增强系统中的对流层精密修正提供了关键技术支撑。