基于FFT和EWT的转子振动信号特征提取方法研究

On the Feature Extraction Method for Rotor Vibration Signal based on FFT and EWT

为了充分挖掘转子振动信号的特征信息,研究提出了一种基于FFT和EWT的转子振动信号特征提取方法。从转子实验台获取转子四种状态的振动信号,将转子特征频率和EWT模态分量组合构成多维特征向量,利用K均值聚类法对比不同方案识别转子状态的正确率,选出最优的特征向量方案,达到了较高的状态识别正确率。

In order to get the full features of the rotor vibration signal,a feature extraction method is proposed based on the fast Fourier transformation(FFT)and the empirical wavelet transform(EWT).The vibration signals of four states of the rotor are obtained from the rotor tests.Multi-dimensional eigenvectors are then composed with the rotor characteristic frequencies and the EWT modal components.The K-means clustering method is used to compare the accuracy of different schemes for identifying the rotor state,and the optimal eigenvector scheme is selected.The results show that high accuracy is achieved in the state recognition with the proposed method.