基于频带熵的滚动轴承故障诊断研究

Rolling element bearing fault diagnosis based on frequency band entropy

发生早期故障的轴承,由于故障特征频率处的能量比较微弱,易于被噪声淹没。通过对滚动轴承故障原理的研究,讨论了利用共振频率解调故障特征频率对于噪声的抑制能力。为了寻找轴承的共振频率,结合时频分析和信息熵的特点,提出了一种新的频带熵方法。首先利用短时傅里叶变换得到信号的时频分布,再沿时间轴计算各个频率的幅值谱熵,从而得到各个频率成分随时间变化的复杂度指标。利用频带熵对轴承共振频率的检测能力,揭示了频带熵与自适应滤波器参数的关系。与原始包络相比,滤波后的包络信号能有效的从噪声中提取出轴承故障特征频率。最后通过仿真和实验验证了该方法的有效性。

Bearing fault characteristic frequencies contain rather weak energy when the early fault occurs,and they are often overwhelmed in noise.According to the research results on the fault mechanism of rolling bearings,the demodulation of bearing resonance frequency is helpful in improving SNR and extracting the characteristic frequencies. Based on time-frequency analysis and information entropy,frequency band entropy （FBE ）was introduced to find the bearing resonance frequency.Short time fourier transform （STFT）was utilized to calculate the time-frequency distribution of signal.Then the amplitude spectrum entropy of each frequency was estimated to discover the complexity of each frequency changing with time.This property provides a way of blindly designing of optimal filter parameters.The bearing characteristic frequencies can be extracted effectively from the signal after filtering.The simulation and experiment verify the validity of the proposed method.