基于小波分形和一类辨识的航空发动机故障诊断

Aeroengine fault diagnoisis based on one-class classification and wavlet-fractal

关键词:  航空发动机  一类辨识+  支持向量机+  小波分形+  故障诊断

Key words:  Aeroengine  One-class identification+  SVM+  Wavlet-fractal  Fault diagnosis

摘要:

在支持向量机理论的基础上,针对支持向量机的二类辨识传统,引入了基于支持向量机的一类辨识理论。设计了航空发动机几种典型故障的一类分类器,使得发动机的故障诊断更加简单可行。同时,将小波分形方法引入到航空发动机振动信号的特征提取中。通过对航空发动机典型故障的成功诊断,证明了该方法的有效性。

Abstract:

Based on SVM(support vector machines) theory,one-class identification theory was introduced.Several classification models,which make the aeroengine fault diagnosis become more simple and viable,were designed based on one-class identification theory.In addition,wavlet-fractal was used to extract feature of aeroengine vibration data.Successful application has been achieved to detect several typical fault of aeroengine.The results show that the one class identification can provide a new effective technology to reveal fault of aeroengine.