基于小波包分析的机械故障特征提取方法研究

Study on the extraction method of mechanical fault symptom based on wavelet packet analysis

研究了一种基于小波包分析与信号能量分解的故障特征提取方法 ,运用这种方法提取了一风机轴不对中故障特征向量 ,为神经网络故障诊断提供了新的故障样本。实验结果表明这种方法比基于Fourier变换的故障特征提取方法更有效 。

A extraction method of mechanical fault symptom bas ed on wavelet packet analysis and signal energy decomposition are researched With this method,an ei ge nvector of shaft-misalignment fault is extracted from ventilator,which provides new fault samples for neural network fault diagnosis The experimental result sh o ws this method is more effective than the extraction method of fault symptom bas ed on the Fourier transformation,and it is very fit for mechanical fault diagnos is