中国空间科学技术

›› 2011, Vol. 31 ›› Issue (6): 8-14.doi: 10.3780/j.issn.1000-758X.2011.06.002

• 研究探讨 • 上一篇    下一篇

雷达测量误差模型的容错检验算法

 胡绍林1,2, 余慧1, 王晓峰2, 楼琳1   

  1. （1 西安卫星测控中心宇航动力学国家重点实验室， 西安 710043） 
    （2 西安理工大学自动化学院， 西安710048）
  • 收稿日期:2010-12-14 修回日期:2011-12-25 出版日期:2011-12-25 发布日期:2011-12-25
  • 作者简介:胡绍林 1964年生,2000年获西安交通大学控制科学与工程专业博士学位，博士后，研究员。研究方向为航天安全与数据处理、故障诊断与容错计算。
  • 基金资助:

    国家自然科学基金 （11026224, 61074077）资助项目

Outliers-tolerant Test Algorithms for Error Models of Radar Measurement

 HU  Shao-Lin1,2, YU  Hui1, WANG  Xiao-Feng2, LOU  Lin1   

  1. （1 State Key Laboratory of Astronautic Dynamics, Xi′an Satellite Control Center, Xi′an   710043）
    （2 School of Automation, Xi′an University of Technology， Xi′an 710048）
  • Received:2010-12-14 Revised:2011-12-25 Published:2011-12-25 Online:2011-12-25

PDF (PC)

1335

被引次数

8

摘要： 空间飞行器导航定位的精度及可靠性很大程度上取决于雷达误差修正模型准确与否。以脉冲雷达为研究对象，采用统计假设检验理论，对雷达测量数据有无野值等不同情形，分别建立了误差模型拟合优度F 检验与容错F-检验、模型分量显著性T-检验和容错T-检验算法,并且利用雷达跟踪飞行器的实测数据对算法进行了验证。结果显示，模型拟合优度检验与模型分量显著性检验等系列算法具有良好的容错性和适应性，可用于诊断雷达误差模型是否正确有效。

关键词: 脉冲雷达, 假设检验, 模型诊断, 容错, 导航, 跟踪

Abstract: The precision and  reliability of spacecraft navigation mainly relies on the accuracy of error models of the radar system and measurement data. For the pulse radar, a series of new test algorithms were built, including  the F- test algorithm and the outliers- tolerant F- test algorithm to test the fittness of the error models with measurement data, and the T- test algorithm and outliers- tolerant T- test algorithm to diagnose the significance of error components. These algorithms were based on the theory of statistical hypothesis test and can be used in two different cases with or without outliers in measurement data respectively. These algorithms were tested with practical measurement data from the pulse radar in tracking a spacecraft. The results show that these algorithms have good outliers tolerance and adaptation, and are available to diagnose whether the error models of radar data are correct.

Key words: Pulse radar, Hypothesis test, Model diagnosis, Outliers tolerance, NavigationTracking