论文部分内容阅读
针对非线性模拟电路故障诊断中参数型故障元件定位的难题,基于Volterra频域核的频谱比较,提出了利用小波滤波器组结合相关分析定位故障元件的方法.首先选择特定的激励信号测定Volterra频域核的非参数频谱;然后用小波滤波器组对得到的正常电路和故障电路的频谱序列进行子带分解;通过计算子带响应序列的相干函数,对正常电路和故障电路进行相关分析,实现参数型故障元件的特征提取.对比实验结果表明,该方法能有效提取故障特征,提高了故障诊断效果.
Aiming at the problem of parametric fault location in fault diagnosis of nonlinear analog circuits, based on the spectral comparison of Volterra frequency domain kernel, a method of locating faulty components by using wavelet filter bank and correlation analysis is proposed. Firstly, the specific excitation signal is selected to measure Volterra frequency Then the wavelet filter bank is used to subband decompose the frequency spectrum of the normal circuit and the fault circuit. The coherence function of the sub-band response sequence is used to analyze the normal circuit and the fault circuit, The feature extraction of parametric faulty components is compared with the experimental results show that this method can effectively extract the fault features and improve the fault diagnosis.