For the weak component in the early failure of the nuclear main pump, it is easy to be masked by strong faults or overwhelmed by strong noise to cause leakage diagnosis, and in actual working condition measurement, multiple sensors are usually used to synchronize the signals. The existing traditional feature analysis methods are only the single-channel vibration signal measured by multi-sensors is processed, and the multi-channel data fusion is not performed at the later stage to achieve the multi-channel synchronization correlation analysis. A multi-dimensional empirical mode decomposition (Multivariate Empirical Mode Decomposition, MEMD)-1.5-dimensional Teager energy spectrum is proposed for the extraction of micro-fault features. Firstly, use the MEMD to adaptively decompose the multi-channel vibration signals on the collected multi-channel fault characteristic signals under the same state to obtain the Intrinsic Mode Functions(IMF) components of each channel, and then calculate the kurtosis value and correlation coefficient of each channel IMF component to select the best IMF component containing the main information of the fault. Finally, the 1.5-dimensional Teager energy spectrum is used to obtain the fault characteristic information of the signal to achieve the extraction of minor fault features. In order to verify the feasibility of the theory, simulation tests are carried out and the method is applied to the early failure of the outer ring of the bearing, and compared with EMD and envelope demodulation, it is verified that this method can effectively deal with early multi-channel failure information of rotating machinery. It has theoretical guidance significance for early diagnosis of small faults of nuclear main pump.
[1]
D. P. Mandic,et al.
Multivariate empirical mode decomposition
,
2010,
Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences.
[2]
Norden E. Huang,et al.
New method for nonlinear and nonstationary time series analysis: empirical mode decomposition and Hilbert spectral analysis
,
2000,
SPIE Defense + Commercial Sensing.
[3]
N. Huang,et al.
The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis
,
1998,
Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences.
[4]
Mohamed Haddar,et al.
Gear fault diagnosis under non-stationary operating mode based on EMD, TKEO, and Shock Detector
,
2019
.
[5]
Qing Zhao,et al.
Maximum correlated Kurtosis deconvolution and application on gear tooth chip fault detection
,
2012
.
[6]
Na Zhao,et al.
Gear fault feature extraction and diagnosis method under different load excitation based on EMD, PSO-SVM and fractal box dimension
,
2019,
Journal of Mechanical Science and Technology.