一种面向卫星共视授时应用的星站钟差融合方法  

A clock bias data fusion method for GNSS common-view time transfer

作　　者：魏吉平 李静怡 王凯旋 唐升[1] WEI Ji-ping;LI Jing-yi;WANG Kai-xuan;TANG Sheng(School of Information Science and Technology,Northwest University,Xi’an 710127,China)

机构地区：[1]西北大学信息科学与技术学院,西安710127

出　　处：《时间频率学报》2023年第2期85-93,共9页Journal of Time and Frequency

基　　金：陕西省自然科学基础研究计划面上资助项目(2021JM303)。

摘　　要：面向GNSS共视授时系统中非连续性多源星站钟差的数据融合需求,提出了一种采用集中式融合结构的数据融合方法。具体地,采用Hampel滤波器剔除钟差观测数据中的奇异值并填补空缺值;采用动态权值分配的加权平均算法实现钟差数据融合;采用Kalman滤波器或α-β滤波器进行跟踪滤波,生成钟差融合终值。仿真实验结果表明,所提出的数据融合方法能够有效抑制观测过程中的误差,输出一组更接近真实值的星站钟差数据,适合星站钟差数据的融合处理,也为其他非连续性多源数据融合应用提供借鉴思路。To meet the data fusion requirement of discontinuous multi-source clock biases(the biases between the local reference time of the stations and the satellite clock)in GNSS common-view time transfer system,a data fusion method using a centralized fusion structure is proposed.Specifically,the Hampel filtering algorithm is used to eliminate the singular values in the observation data and fill in some missing values of clock biases;The clock bias fusion is realized using the weighted average algorithm of dynamic weight distribution;Kalman filter orα-βfiltering algorithm is used for tracking filtering to generate the final value of clock bias fusion.The simulation experiment results show that the data fusion method in this paper can effectively suppress the error in the observation process and output a set of biases between the stations and the satellite close to the true value,which is suitable for the fusion processing of satellite clock biases data,and also provide referring ideas for other discontinuous multi-source data fusion applications.

关 键 词：共视授时 数据融合 Hampel滤波 跟踪滤波 动态权值分配 

分 类 号：P228.4[天文地球—大地测量学与测量工程]