基于小波变换和粗糙集的电力电子电路故障诊断方法

A Fault Diagnosis Method for Power Electronic Circuits Based on Wavelet Transformation and Rough Set

为了快速而有效地诊断电力电子电路的故障,提出了一种小波变换与粗糙集相结合的新方法。用小波变换分解输出电压信号以便提取故障特征信息,在此基础上,构建了条件属性集与决策属性集,运用了粗糙集的数据挖掘能力去除冗余条件属性,约简后提取出故障诊断规则,按照这些规则来诊断各类故障。最后,通过数学建模与数据处理,仿真结果验证了快速性和有效性。

For a quick and effective diagnosis of faults in power electronic circuits, this paper presents a new approach combining wavelet transformation with rough set, wherein the wavelet transformation decomposes the output voltage signal so as to extract information about fault characteristics. Condition attribute set and decision attribute set are created on that basis, the data mining ability of the rough set is used to remove redundant condition attributes, and rules for fault diagnosis are extracted after reduction. Finally, through mathematic modeling and data processing, the simulation result verifies the quickness and effectiveness of the new approach.