基于小波变换预处理的模拟电路故障诊断

把小波变换用于模拟电路故障诊断中。输出信号的低频系数反映了信号的概貌特征．表征了电路的故障情况,因此,利用小波变换有效地提取故障特征信息。提出了“小波系数-故障”的故障诊断方法。计算和实验结果表明：该方法可以简化神经网络的结构和减少它的训练时间,快速高效地进行模拟电路故障诊断与定位。

Fault diagnosis of analog circuit based on wavelet transform preconditioning

In this paper, the wavelet transform is put into use of fault diagnosis of analog circuit.The low frequency coefficient of output signal shows fault of analog circuit,so,it abstracts the fault character information by using wavelet transform effectively.A new method about fault diagnosis "Wavelet coefficient - Fault" is referred.Simulati on experiments prove that the proposed method is available and gives better performance and can simplifying neural network architec ture and minimizing its training and processing time.