基于小波包能量与峭度谱的滚动轴承故障诊断

Fault diagnosis of rolling bearing based on wavelet packet energy and spectral kurtosis

针对故障轴承的振动信号中包含冲击成分,导致信号的能量集中的问题,提出了一种基于小波包能量与峭度谱相结合的方法用以提取轴承故障信号特征。首先应用小波包对测量信号进行分解、能量归一化处理和信号重构,然后将重构信号采用峭度谱确定带通滤波器的最佳中心频率和带宽,最后将滤波信号进行包络解调并提取故障特征频率。分别对仿真信号和试验数据进行研究,能够清晰地得到故障特征频率及其高次谐波,从而验证了所提方法在滚动轴承的故障诊断中的有效性和可行性。

Since the impact component contained in the vibration signal of fault bearing leads to energy concentration of the signal,the method based on wavelet packet energy and spectral kurtosis is proposed in this article,in order to extract the characteristics of the related signal. Firstly,the vibration signal of rolling bearing is decomposed;the energy is calculated and reconstructed by means of the wavelet packet. Then,the best frequency and bandwidth of the bandpass filter are determined automatically by means of spectral kurtosis. Finally,the filtered signal is subjected to envelope analysis,in order to extract the fault’s characteristic frequency. Based on the simulation signals and experimental data,the fault’s characteristic frequency and its higher harmonics are clearly obtained,thereby verifying the effectiveness and feasibility of the proposed method for fault diagnosis of rolling bearing.