基于峭度特征提取的模拟电路故障诊断

Analog Circuit Fault Diagnosis Based on Kurtosis Feature Extraction

针对非线性容差模拟电路中存在故障特征提取难的问题,结合小波分析技术理论,引用了峭度来描述的故障信息的方法进行故障特征提取,并且利用领域覆盖算法作为分类器,进行故障诊断。仿真实验结果表明,峭度能较好的应用于模拟电路故障诊断。

To slove the problem of the difficulty in extracting the feature of the fault in tolerance analog circuits,this paper which based on the theory of wavelet analysis uses kurtosis to extract the fault information,it also uses the neighborhood covering algorithm as a classifier to diagnose the fault.The simulation result shows that kurtosis is ap-plied in diagnosing the fault of analog circuits.