BDS/GNSS实时卫星钟差估计多模融合质量控制方法研究

The precision and reliability of the real-time satellite clock estimation is directly determined by the data quality control, which is also the key to guarantee real-time precise positioning service. The combination of the BeiDou navigation satellite system (BDS) and other GNSS systems not only increases the strength of the common parameters, but also improves the precision of the clock solution. At present, since the quality control is still imperfect for the real-time combined clock estimation, it’s hard to make sure that the multi-GNSS real-time clock estimation is reliable and stable. Therefore, three core technologies, which include the priori quality control of the combined clock estimation, the dynamic quality control during the estimation process and the posterior quality control of ambiguity parameters, will be studied. Firstly, the BDS/GNSS combined estimation model will be refined and the priori recognition model will be built to control the anomaly of parameter estimation. Secondly, aiming at solving the instability of the real-time clock estimation quality and the anomaly of some satellite clock offsets, the adaptively quality control method of the real-time combined clock estimation will be researched. Meanwhile, the anomaly will be classified to optimize the estimation and the stability control algorithm will be designed for the estimation process. Thirdly, the ambiguity resolution method in BDS real-time clock estimation will be proposed for the posterior quality control of float ambiguity parameters. Finally, the quality control algorithm software will be developed for the real-time combined clock estimation and some experiments will be carried out. In conclusion, the prospective researches in this project are able to provide the technology support for our BDS real-time precise satellite clock offsets service.

数据质量控制直接决定了实时卫星钟差估计的精度和可靠性，也是保障实时精密位置服务的关键。北斗卫星导航系统（BDS）与其他GNSS的融合处理可增强共同参数的强度，提高钟差解算精度，但目前针对钟差估计多模融合的质量控制方法仍不完善，难以保障可靠稳定的实时多模钟差估计。课题针对实时钟差估计的多模融合先验质量控制、估计过程动态质量控制、验后模糊度参数质量控制三个核心技术展开攻关。首先精化BDS/GNSS融合估计模型和建立先验识别模型控制参数估计异常；其次针对实时钟差估计质量不稳定及个别卫星钟差异常问题，研究适合于实时多模融合钟差估计的自适应动态质量控制方法，构建估计异常分类优化和估计过程稳定控制算法；第三针对浮点模糊度参数进行验后质量控制问题，攻关BDS实时钟差估计中模糊度在线固定方法；最后研制一套实时钟差估计多模融合质量控制算法软件并开展实验验证，为我国的BDS实时卫星精密钟差服务提供技术支撑。

卫星钟差是实时精密单点定位（PPP）中的一项重要误差改正项，实时质量控制对于保障可靠的GNSS高精度实时钟差估计至关重要。本项目针对系统间偏差（ISB）模型精化、实时质量控制算法、验后模糊度固定等内容，开展了BDS/GNSS实时卫星钟差估计质量控制与性能提升方法研究。取得的主要研究成果如下：（1）针对BDS-2和BDS-3卫星在不同类型接收机上码硬件延迟偏差不一致的现象，研究了BDS-2和BDS-3实时钟差估计的融合策略，提出了ISB分段常数处理算法。结果表明，相对于ISB白噪声方案，分段常数处理将BDS-2 MEO卫星钟差估计精度提升了42.3%，相对于不引入ISB的情况，BDS-3提升18.5%。（2）针对钟差跳变问题，提出了基准自动平稳切换算法，提高了钟差估计精度和连续光滑性。针对GNSS观测值容易受到异常值影响并导致数据质量变差的问题，提出了一种基于改进四分位距的实时卫星钟差估计质量控制方法，与基于中位数的质量控制方法相比，该方法估计的卫星钟差更精密。（3）针对钟差重收敛时间长难题和继续提升精度需求，提出了顾及在线相位小数偏差（FCB）滤波的整数钟实时估计和钟差瞬时重收敛算法。结果表明，所提方法整数钟精度达0.04ns，提升了约31.0%，不仅达到IGS分析中心的顶级性能水平，且解决了钟差重收敛问题，能直接实现PPP模糊度固定。同时，还增加了在GNSS/低轨卫星增强定位中的应用研究，服务于BDS精密定位服务应用推广。