航空学报 > 2005, Vol. 26 Issue (6): 743-747

一种联邦卡尔曼滤波渐变型故障容错方法

张国良1, 曾静2, 李陶1, 邓方林1   

  1. 1. 第二炮兵工程学院301室, 陕西 西安 710025;2. 第二炮兵工程学院 数学室, 陕西 西安 710025
  • 收稿日期:2004-09-06 修回日期:2005-03-09 出版日期:2005-12-25 发布日期:2005-12-25

Fault-tolerance Designing of Federal Kalman Filter for Gradual Changing Fault

ZHANG Guo-liang1, ZENG Jing2, LI Tao1, DENG Fang-lin1   

  1. 1. Faculty 301, The Second Artillery Engineering College, Xi'an 710025, China;2. Faculty of Mathematics, The Second Artillery Engineering College, Xi'an 710025, China
  • Received:2004-09-06 Revised:2005-03-09 Online:2005-12-25 Published:2005-12-25

摘要: 具有良好的容错性是组合导航系统的基本要求。组合导航系统易于完成突变型故障的检测与容错,对渐变型故障的检测与容错则较为困难。文中以联邦卡尔曼滤波为组合导航系统状态估计方法,针对渐变型故障,提出了观测品质的概念及其模糊评估方法,应用观测品质进行了导航系统的渐变型故障检测,利用观测品质构成局部滤波器估计均方误差阵的修正矩阵,设计了组合导航系统智能自适应渐变型故障容错方法并完成了仿真运算。

关键词: 容错滤波, 观测品质, 组合导航, 联邦卡尔曼滤波

Abstract: It is fundamental for integrated navigation system to possess fine fault tolerance. It is relatively easy for integrated navigation system to accomplish the abrupt changing fault detection and fault tolerating, as for gradual changing fault it is rather difficult. The federal Kalman Filter is applied in this essay for the state estimation of integrated navigation system, and the concept of observation quality as well as its fuzzy evaluation is presented to deal with the gradual changing fault. The observation quality is applied to detect the gradual changing fault of navigation system, and the observation quality is used to destruct the modified matrix of mean square error matrix of local filter estimation, the intelligent adaptive fault tolerance for gradual changing fault is designed for integrated navigation system. Finally, the simulation is accomplished.

Key words: fault-tolerance filter, observation quality, integrated navigation, Federal Kalman filter

中图分类号: