车桥振动噪声信号特征提取方法的研究

Study on Feature Extraction of Drive Axle Vibration or Noise Signals

从相关性原理出发给出了一种对两路非稳态随机振动噪声信号进行解耦的算法,通过对人工合成的两路信号进行解耦验证了此算法的正确性。利用小波包对解耦后的信号进行分解,提取其各频段的能量值,通过其统计量作为信号的特征值。并将这种方法应用到车桥的振动噪声信号的特征提取上,取得了良好的效果。

According to correlation theory, the paper gives a algorithm which can be used to decoupling two non-stationary random vibration signals or noise signals, the algorithm is proved to be correct by decoupiing two artificial signals, the decoupled signals are decomposed by using wavelet packet and every frequency bands＇ energies are extracted. Statistical values are used to be the signals＇ features, the method is applied into the extracting of drive axle＇s vibration or noise features.