中国空间科学技术

中国空间科学技术

• •    

基于神经网络模型的层次诊断方法及应用

程惠涛,黄文虎,姜兴渭   

  1. 哈尔滨工业大学
  • 发布日期:1998-04-25

NEURAL NETWORK BASED HIERARCH DIAGNOSIS METHOD AND ITS APPLICATION

Cheng Huitao Huang Wenhu Jiang Xingwei (Harbin Institute of Technology,Harbin 150001)   

  • Online:1998-04-25

被引次数

1

摘要： 对神经网络和故障诊断的层次分解模型进行了研究，把神经网络技术和层次分解模型相结合，利用神经网络来实现分层诊断。以卫星能源系统为背景，建立了该系统故障模拟试验台的层次分解模型，在此基础上构造了合适的分层神经网络，通过分层神经网络诊断系统故障，给出的两个实例说明了这种方法的有效性。

关键词: 卫星, 故障诊断, 层次分解模型, 人工神经元网络

Abstract: Neural network and hierarch classification method in fault diagnosis are studied.Neural network technology connected with hierarch classification model is to rea lize hierarch diagnosis.Satellite energy sources is taken for the research object,and the hierarch classification model of fault analogue test bed of this system is established.Based upon this,proper hierarch neural network is established and system fault is diagnosed.This method is verified and validated by two examples.