针对轮轨激扰下高速动车组轴箱轴承故障信息提取困难的问题,提出了一种时频能量谱尺度空间表征(Time Frequency Energy Spectrum Scale Space Representation,TFESSSR)的故障诊断方法.该方法首先通过短时傅里叶变换(Short Time Fourier Transform,STFT)计算轴箱振动信号的二维时频表征,并对二维时频表征沿时间轴进行积分得到时频能量谱(Time Frequency Energy Spec-trum,TFES);然后使用尺度空间表征方法(Scale Space Representation,SSR)从振动信号的时频能量谱中识别出轴承故障冲击的共振频带边界;最后通过经验小波变换(Empirical Wavelet Transform,EWT)根据所识别的共振频带边界设计滤波器组对轴箱振动信号进行分解.使用高速动车组滚动振动试验台试验数据对算法有效性进行了验证,并与传统尺度空间表征算法进行了对比.结果表明,相较于传统的SSR方法,TFESSSR方法能够在轮轨激扰下更快速、更准确识别轴箱轴承故障冲击共振频带边界,实现轴箱轴承的故障诊断.
Axle box Bearing Fault Diagnosis Method for High-speed Multiple Units Based on TFESSSR
To solve the problem of difficult extraction of axle box bearing fault information of high-speed multiple unit under wheel-rail excitation,the article proposes a fault diagnosis algorithm with Time Frequency Energy Spectrum Scale Space Representation(TFESSSR).This method firstly calculates the bidimensional time frequency representation of vibration signals of the axle box through Short Time Fourier Transform(STFT)and integrate the bidimensional time frequency representation along the time axis to obtain the Time Frequency Energy Spectrum(TFES);then it uses the Scale Space Representation(SSR)method to identify the resonance frequency band boundary of bearing fault impact from the TFES of the vibration signals;finally,using the Empirical Wavelet Transform(EWT),it designs the filter group according to the identified resonance frequency band boundary to decompose the vibration signals of the axle box.It uses the test data got from the rolling and vibration test bench for high-speed multiple unit to verify the effectiveness of the algorithm and compares it with the conventional SSR algorithm.The result shows that compared with the conventional SSR method,the TFESSSR method can identify the resonance frequency band boundary of the axle box bearing fault impact more quickly and accurately under wheel-rail excitation so as to realize the fault diagnosis of axle box bearing.
Scale Space Representation(SSR)Time Frequency Energy Spectrum(TFES)axle box bearingfault diagnosiswheel-rail excitation
NETL
NSTL
万方数据
