基于时频幅值匹配的列车轴承故障声学信号的多普勒校正方法

摘要

由于列车运动所产生的多普勒频移和扩展现象，给轴承的故障识别、诊断带来了难度。提出了一种基于时频幅值匹配的多普勒校正方法，并将其运用到列车轴承上。首先，依据莫尔斯声学理论和短时傅里叶变换，推导出时频幅值曲线与中心时间的关系，通过匹配理论与实际信号的时频幅值曲线求出相对应的中心时间。而后，依据该中心时间建立重采样时间序列，并以此实现信号的多普勒校正。通过仿真信号和实际信号验证了该方法的有效性。%The Doppler frequency shift and extension phenomenon bring serious difficulty to the bearing fault identification and diagnosis.A Doppler correction method was proposed based on the time-frequency amplitude matching and applied to the bearing fault identification.On the basis of the Morse acoustic theory and short-time Fourier transform (STFT),the relationship between the time-frequency amplitude and center time was concluded.The optimal center time was acquired by the best matching between the theoretical and practical time-frequency amplitudes to establish resampling time series and remove the Doppler distortion.The effectiveness of the method was demonstrated by using simulated signals and experimental signals.