首页 | 本学科首页   官方微博 | 高级检索  
     

基于Volterra频域核辨识的非线性模拟电路故障诊断
引用本文:刘本德,胡昌华.基于Volterra频域核辨识的非线性模拟电路故障诊断[J].控制与决策,2009,24(8).
作者姓名:刘本德  胡昌华
作者单位:第二炮兵工程学院302教研室,西安,710025
基金项目:国家自然科学基金重点项目,教育部新世纪优秀人才支持计划项目 
摘    要:基于Volterra级数时域频域混合模型,提出了辨识非线性模拟电路频域核的故障诊断方法.利用混合模型辨识算法和范德蒙特法估计各种故障状态下电路响应的前3阶频域核,提取故障特征并与相应的故障模式一起构成特征样本集,借助于支持向量机多分类器进行分类识别,实现非线性模拟电路的故障诊断.阐述了诊断原理及诊断步骤,并给出了诊断实例.仿真结果表明,该方法的故障识别率较高,便于计算机计算.

关 键 词:非线性电路  故障诊断  Volterra级数  频域核  支持向量机

Fault diagnosis of nonlinear analog circuits based on Volterra frequency domain kernel identification
LIU Ben-de,HU Chang-hua.Fault diagnosis of nonlinear analog circuits based on Volterra frequency domain kernel identification[J].Control and Decision,2009,24(8).
Authors:LIU Ben-de  HU Chang-hua
Affiliation:302 Unit;The Second Artillery Engineering Institute;Xi'an 710025;China.
Abstract:Based on Volterra time/frequency domain algorithm,a method of fault diagnosis is presented,which can identify frenquency kernel of nonlinear analog circuit.The forward three order Volterra kernels in frequency domain can be obtained by using the identifying algorithm and Vandermonde method under faulty models.Circuits fault features are extracted,which are combined with faulty modes and regarded as faulty characteristic samples.The support vector machine multi-classifier is trained by faulty characteristic ...
Keywords:Nonlinear circuit  Fault diagnosis  Volterra series  Frequency domain kernels  Support vector machine  
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号