Scholarlink

<p>The International GNSS Service (IGS) requires advanced multi-GNSS orbit combination strategies to replace current GPS/GLONASS-focused operations with consistent products covering GPS, GLONASS, Galileo, and BDS. We developed an enhanced orbit combination methodology using a modified Förstner Variance Component Estimation (VCE) scheme that optimizes weighting strategies through data clustering approaches, including individual satellite weighting, satellite-type grouping, and machine-learning-generated clusters. Our novel approach incorporates a priori knowledge from Satellite Laser Ranging (SLR) orbit validations and sequential weight information from previous combinations to refine Analysis Center (AC) weights. Sequential weight estimation significantly reduces day boundary orbit misclosures and stabilizes temporal AC weight variability. The combined solutions demonstrate exceptional inter-consistency with RMS values below 3–5&#xa0;mm for GPS and Galileo, while GLONASS and BDS show higher variability (10–15&#xa0;mm), highlighting the importance of satellite grouping strategies. Intermediate grouping approaches based on IGS metadata or hierarchical clustering provide optimal balance between constellation-level oversimplification and satellite-specific day-to-day variability. SLR-based knowledge incorporation offers targeted improvements, particularly for challenging high and low β angle conditions, demonstrating the effectiveness of external validation in multi-GNSS orbit combination.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Advancing multi-GNSS orbit combination in the variance component estimation framework

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月30日

查看全文
  • Radosław Zajdel,
  • Gustavo Mansur,
  • Pierre Sakic,
  • Paul Rebischung,
  • Andreas Brack,
  • Benjamin Männel,
  • Jan Douša

摘要

The International GNSS Service (IGS) requires advanced multi-GNSS orbit combination strategies to replace current GPS/GLONASS-focused operations with consistent products covering GPS, GLONASS, Galileo, and BDS. We developed an enhanced orbit combination methodology using a modified Förstner Variance Component Estimation (VCE) scheme that optimizes weighting strategies through data clustering approaches, including individual satellite weighting, satellite-type grouping, and machine-learning-generated clusters. Our novel approach incorporates a priori knowledge from Satellite Laser Ranging (SLR) orbit validations and sequential weight information from previous combinations to refine Analysis Center (AC) weights. Sequential weight estimation significantly reduces day boundary orbit misclosures and stabilizes temporal AC weight variability. The combined solutions demonstrate exceptional inter-consistency with RMS values below 3–5 mm for GPS and Galileo, while GLONASS and BDS show higher variability (10–15 mm), highlighting the importance of satellite grouping strategies. Intermediate grouping approaches based on IGS metadata or hierarchical clustering provide optimal balance between constellation-level oversimplification and satellite-specific day-to-day variability. SLR-based knowledge incorporation offers targeted improvements, particularly for challenging high and low β angle conditions, demonstrating the effectiveness of external validation in multi-GNSS orbit combination.