基于小波包分形的电力电子故障诊断方法

Fault Diagnosis Method of Power Electronic Circuits Based on Wavelet Package Fractal Analysis

针对现有电力电子故障诊断的问题,提出了一种基于小波包分形的电力电子故障的诊断方法。首先使用3层小波包对电路输出电压波形进行分解、消噪和重构等预处理,以获得不同频带的重构信号,然后再计算各个重构信号的分形维度,提取出相应的故障特征;最后以此作为输入数据来训练支持向量机,实现电力电子故障的诊断。以三相整流电路为例进行实验,实验结果表明该方法能有效地提取电力电子故障诊断的故障特征,提高故障诊断的准确率,可有效地应用于电力电子系统的故障诊断。

In order to solve the problem of power electronic circuits fault diagnosis, a diagnosis method of power electronic circuits fault based on wavelet package fractal analysis is proposed. Firstly, the circuit output voltage waveforms are decomposed by three-level wavelet package, after de-noising and reconstructive process, the re- constructed signals of different frequency bands are obtained. Then the fractal dimension of each reconstructed signal is calculated, and the corresponding fault feature is extacted. Finally, these fault vectors are taken as the input data to train the support vector machine （SVM）, which achieves the diagnosis of power electronic circuit fault. The experimental result shows that the proposed method can effectively extract the essential features of fault output signals and improve the accuracy of fault diagnosis, which can be applied to the fault diagnosis of power electronic circuits effectively.