基于小波包-支持向量机的发动机泵机组故障诊断

Fault Diagnose of Engine Pump Machine Based on Wavelet and Support Vector Machine

发动机泵机组由发动机、油泵以及增速器构成,泵机组故障的振动信号往往淹没在发动机的巨大振动中,对其进行故障诊断非常困难。本文利用频谱相加降噪以及小波多分辨率特性成功地将泵机组故障信号提取出来,并利用小波包与支持向量机理论相结合的方法,对泵机组的故障模式进行识别,取得了很好的效果。

The engine pump machine is composed of engine,oil pump and speed increase implement,the fault vibration signal of pump machine is sometimes submerged in enormous engine vibration,the fault diagnose to it is very difficult.The fault vibration signal of pump machine was extracted out successfully with denoising of frequency spectrum addition and variable resolution of wavelet.The fault mode of pump machine was identified making use of wavelet packet and support vector machine,good result was got.