English

应用科学学报 ›› 2010, Vol. 28 ›› Issue (1): 72-76.

• 控制与系统 • 上一篇    下一篇

应用自组织模糊神经网络估计卫星姿态系统执行机构故障

程月华1;2， 姜斌1， 杨明凯1， 高志峰1   

  1. 1. 南京航空航天大学自动化学院，南京210016
    2. 南京航空航天大学高新技术研究院，南京210016
  • 收稿日期:2009-06-01 修回日期:2009-10-22 出版日期:2010-01-20 发布日期:2010-01-20
  • 作者简介:程月华，博士生，助理研究员， 研究方向：小卫星姿态控制、故障诊断与容错控制，E-mail：chengyuehua@ nuaa.edu.cn；姜斌，博士，教授，博导，研究方向：故障诊断与容错控制，E-mail：binjiang@nuaa.edu.cn
  • 基金资助:

    国家自然科学基金(No.9081603)； 南京航空航天大学科研创新团队培育基金(No.Y0847-031)资助

Self-Organizing Fuzzy Neural Network-Based Actuator Fault Estimation for Satellite Attitude Systems

CHENG Yue-hua1;2, JIANG Bin1, YANG Ming-kai1, GAO Zhi-feng1   

  1. 1. College of Automation Engineering, Nanjing University of Aeronautics and Astronautics,
    Nanjing 210016, China
    2. Academy of Frontier Science, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China
    Abstract:
  • Received:2009-06-01 Revised:2009-10-22 Online:2010-01-20 Published:2010-01-20

PDF

1537

被引次数

19

摘要：

自组织模糊神经网络可以根据系统状态在线更新权值和调整节点，优化网络结构. 文中针对某卫星姿态控制系统提出了基于两个自组织模糊神经网络的执行机构故障诊断方法. 网络SOFNN1用于健康系统的离线训练，估计出系统的不确定项和扰动项，网络输出结果作为故障检测的阈值参考. 网络SOFNN2在网络SOFNN1的基础上估计执行器故障. 仿真表明，在噪声干扰和系统参数不确定的情况下，在线自组织模糊神经网络结构的方法能很好地估计系统执行机构故障，比固定结构的模糊神经网络估计速度快，因此更具优越性.

关键词: 自组织模糊神经网络, 故障估计, 卫星, 执行机构

Abstract:

Weights and nodes of a self-organizing fuzzy neural network (SOFNN) can be updated online for network structure optimization. This paper studies a robust fault diagnostic approach based on two SOFNNs for a class of satellite attitude dynamics. The designed SOFNN1 is used to estimate uncertainties and external perturbations of fault-free satellite attitude dynamics, whose output is chosen as a referenced threshold of fault
detection. Based on SOFNN1, SOFNN2 is constructed to estimate actuator faults occurring in the satellite attitude dynamics. Simulation results demonstrate that SOFNN has good dynamics performance in estimating actuator faults for the considered dynamics with external noise and system parameter uncertainties. Compared with fixed-structured FNN, the proposed SOFNN has advantages in estimation speed.

Key words: self-organizing fuzzy neural network, fault estimation, satellite, actuator

中图分类号: