中国空间科学技术

中国空间科学技术 ›› 2022, Vol. 42 ›› Issue (1): 16-29.doi: 10.16708/j.cnki.1000-758X.2022.0002

• 综述 • 上一篇    下一篇

卫星在轨故障案例与人工智能故障诊断

王亚坤，杨凯飞，张婕，郭莉芳，韩笑冬，邢川，王睿   

  1. 中国空间技术研究院 通信与导航卫星总体部，北京100094
  • 出版日期:2022-02-25 发布日期:2022-01-27
  • 基金资助:
    国家自然科学基金面上项目（61972398）；173计划重点项目（2019-JCJQ-ZD-342-00）

Case study of inorbit satellite failures and artificial intelligence based failure detection

WANG Yakun, YANG Kaifei, ZHANG Jie, GUO Lifang, HAN Xiaodong, XING Chuan, WANG Rui    

  1. Institute of Telecommunication and Navigation Satellites, China Academy of Space Technology，Beijing 100094，China
  • Published:2022-02-25 Online:2022-01-27

PDF (PC)

1153

摘要： 减少卫星在轨故障，使卫星按设计寿命要求在轨执行飞行任务是卫星设计的一个重要课题。为了分析在轨卫星故障的原因，探索最新的卫星故障诊断方法，首先收集了2003～2021年卫星典型在轨故障案例，然后对卫星故障从不同的维度进行统计分析和相关性分析，包括卫星平台与卫星故障的相关性、卫星在轨时长与卫星故障的相关性等，最后介绍了基于人工智能的故障诊断研究，包括国内外应用的基于人工智能的卫星故障诊断系统和故障诊断方法，以及针对时间序列数据的人工智能故障诊断方法。本研究可以为今后卫星系统设计、避免故障发生提供指导。

关键词: 在轨卫星, 故障案例, 故障统计分析, 人工智能, 时间序列数据

Abstract: Reducing satellite failures and keeping satellites healthy in orbit are important issues in satellite design. In order to analyze the causes of satellite failures and explore the latest satellite failures detection methods, we first collected typical satellite fault cases in orbit from 2003 to 2021, and then performed statistical analysis and correlation analysis of satellite failures from different aspects, including the correlation between satellite platforms and satellite failures, and the relationship between satellite in-orbit time and satellite failures, etc. Finally, the artificial intelligence based satellite failures detection research was introduced, including the artificial intelligence based failures detection systems and methods as well as the time-series data failures detection research. This research can provide reference for future failure case analysis and failure detection system design of satellite.

Key words: in-orbit satellite, failure cases, failure statistical analysis, artificial intelligence, time-series data