系统工程与电子技术 ›› 2026, Vol. 48 ›› Issue (8): 2833-2840.doi: 10.12305/j.issn.1001-506X.2026.08.28

• 制导、导航与控制 • 上一篇    

基于稳定度加权和Kalman滤波算法相融合的多级时间比对融合算法

宁学友, 刘强, 胡邓华, 张爽   

  1. 空军工程大学防空反导学院,陕西 西安 710051
  • 收稿日期:2025-06-10 修回日期:2025-12-12 出版日期:2026-05-14 发布日期:2026-05-14
  • 通讯作者: 刘强
  • 作者简介:宁学友(1997—),男,硕士研究生,主要研究方向为高精度时间同步技术
    胡邓华(1982—),男,副教授,博士,主要研究方向为数据传输与通信网络仿真
    张 爽(1990—),男,讲师,博士,主要研究方向为高精度时间同步技术
  • 基金资助:
    国家自然科学基金(61701525)资助课题

Multi-level time comparsion fusion algorithm based on the intergration of stability weighting and Kalman filtering algorithm

Xueyou NING, Qiang LIU, Denghua HU, Shuang ZHANG   

  1. Air and Missile Defense College,Air Force Engineering University,Xi’an 710051,China
  • Received:2025-06-10 Revised:2025-12-12 Online:2026-05-14 Published:2026-05-14
  • Contact: Qiang LIU

摘要:

针对卫星双向时间频率传递(two-way satellite time and frequence transfer,TWSTFT)和精密单点定位(precise point positioning,PPP)等单一技术难以同时满足高精度和高稳定度的需求,提出一种多级时间比对融合算法。模型一基于稳定度权重融合TWSTFT、GALILEO P3和全球定位系统(global positioning system,GPS)P3码,再利用Kalman滤波将融合结果与GPS PPP进行二次融合,模型二优先融合TWSTFT与GPS PPP,再通过两级Kalman滤波逐级融合GALILEO P3和GPS P3码。仿真结果表明,两种模型均能改善时间比对的精度和稳定度,模型一适合对稳定度要求较高的长期钟差分析场景,模型二适合实时监测和动态环境下的时间比对需求。数据失效测试进一步表明,模型二在GPS PPP缺失时表现更好,两种模型对GALILEO P3和GPS P3码缺失时均表现出良好的容错性,验证了算法设计的容错能力。

关键词: 时间比对, 数据融合, Kalman滤波, 稳定度加权

Abstract:

For the single techniques such as two-way satellite time and frequence transfer (TWSTFT) and precise point positioning (PPP), which are difficult to satisfy the demands of high accuracy and high stability at the same time, a multi-level time comparison fusion algorithm is proposed. Model 1 fuses TWSTFT, GALILEO P3, and global positioning system (GPS) P3 codes with stability weighting, and then fuses the weighted results with GPS PPP using Kalman filtering, while Model 2 prioritizes the fusion of TWSTFT and GPS PPP, and then fuses GALILEO P3 and GPS P3 codes step by step by two-level Kalman filtering. Simulation results show that both models can improve the accuracy and stability of time comparison, and Model 1 is suitable for long-term clock difference analysis scenarios with high stability requirements, while Model 2 is suitable for real-time monitoring and time comparison needs in dynamic environments. The data failure test further shows that Model 2 performs better when GPS PPP is missing, and both models show good fault tolerance for GALILEO P3 and GPS P3 code missing, which verifies the fault tolerance of the algorithm design.

Key words: time comparison, data fusion, Kalman filtering, stability weighting

中图分类号: