期刊论文

A Novel Method of Analog Circuit Fault Diagnosis

中文

怀化学院学报

1671-9743

2017

36

11

47-51

10.16074/j.cnki.cn43-1394/z.2017.11.012

16C0637:湖南省教育厅科学研究项目 E51366:湖南科技大学校级课题 61202462:国家自然科学基金 61308001:国家自然科学基金

本校为其他机构

为了更准确的诊断模拟电路故障,并节约诊断时间,提出了提取故障信号的峭度和偏度作为特征向量的故障诊断方法,相对于传统的以二阶统计量为基础的故障特征提取方法,峭度和偏度能有效地抑制噪声影响,显著减少了信号的失真,从而最大限度地保持了故障信息的原貌;并对信息融合技术加以运用,不再单一的运用电压信号作为构造故障特征向量的信息源,而是同时采样电压和电流信号作为故障特征提取的信息源,增加了故障信号的有效信息量.通过对实例电路的仿真诊断结果表明,该方法故障诊断率高、诊断时间短,不失为一种新的有效的模拟电路故障诊断方法.

In order to diagnose analog circuit fault more accurately, and save the diagnostic time, proposed the method of extraction fault signal of kurtosis and skewness as features vector. Compared with traditional method of fault feature extraction based on second order statistics, kurtosis and skewness can suppress noise effectively, reduce the distortion of signal significantly, and maximize keep the original appearance of fault information; and used information fusion technology, not limited to extraction single of voltage signal as the source of fault feature vector, but sampled signal of voltage...