节点文献

基于解轨迹多项式系数估计的非线性电路故障诊断

Fault diagnosis for weakly nonlinear circuits based on polynomial coefficient estimation of trajectory

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 徐晨曦陈光

【Author】 Xu Chenxi~(1,2) Chen Guangju~1~1(CAT Lab of school of automatic Engineering of UESTC,Chengdu 610054,China)~2(Dep.of Radar system engineering of Air force radar academy,Wuhan 430019,China)

【机构】 电子科技大学自动化工程学院电子科技大学自动化工程学院 成都610054空军雷达学院雷达系统工程系武汉430019成都610054

【摘要】 本文针对一类斜波激励响应轨迹(解轨迹)存在的弱非线性电路,提出一种基于解轨迹多项式拟合的非线性电路故障诊断方法。本文在电路混合节点方程基础上,通过解轨迹多项式系数迭代关系分析多项式系数与故障信息间的对应关系;由解轨迹多项系数间存在的由低到高的解耦关系导出基于多项式系数的故障诊断方程,将非线性电路故障诊断转换为一组线性电路故障诊断问题。以解轨迹多项式系数为特征的故障诊断方法能有效减少故障传播的非线性造成的故障特征重叠、交叉现象,并可充分利用电路结构信息实现故障隔离;文中同时给出了基于解轨迹多项式系数聚类性分析的测试点优化准则。该方法具有分析过程简单、分析动态范围大和适用于未知非线性电路故障诊断等特点。最后通过实例仿真说明该方法的有效性。

【Abstract】 In this paper,an approach of faults diagnosis is presented for weakly nonlinear circuits which exists a trajectory produced by ramping power supply.With Taylor series expansion,the trajectory can be represented by a polynomial approximately,and the iteration expressions of the polynomial’s coefficients from low to high in order is derived from the Taylor series of the trajectory.Then,the relationship of between the polynomial’s coefficients and faults information is analyzed,and the uncoupling relationship of between the coefficients is created in rising order.Based these relationships,a group of linear diagnosis equations are structured in rising order,that is,the problem of nonlinear diagnosis is transformed to linear diagnosis problems.In proposed approach,the knowledge of circuit structure can be used for faults isolation if the CUT(circuit under test) is known,and the crosswise and overlap of between different faults characteristics caused by nonlinear spread of faults effect can be effectively eliminate with the trajectory fitting of a polynomial.Thus,the proposed is beneficial for nonlinear faults diagnosis whether the CUT is known or unknown.Otherwise,an optimization way of test-points is also discussed based on polynomial’s coefficients in here.Finally,two examples are given to illuminate the proposed is available.

【基金】 国家自然科学基金(60372001)资助项目
  • 【文献出处】 仪器仪表学报 ,Chinese Journal of Scientific Instrument , 编辑部邮箱 ,2006年10期
  • 【分类号】TM132
  • 【被引频次】3
  • 【下载频次】79
节点文献中: 

本文链接的文献网络图示:

本文的引文网络