Machine fault feature extraction based on intrinsic mode functions

This work employs empirical mode decomposition (EMD) to decompose raw vibration signals into intrinsic mode functions (IMFs) that represent the oscillatory modes generated by the components that make up the mechanical systems generating the vibration signals. The motivation here is to develop vibration signal analysis programs that are self-adaptive and that can detect machine faults at the earliest onset of deterioration. The change in velocity of the amplitude of some IMFs over a particular unit time will increase when the vibration is stimulated by a component fault. Therefore, the amplitude acceleration energy in the intrinsic mode functions is proposed as an indicator of the impulsive features that are often associated with mechanical component faults. The periodicity of the amplitude acceleration energy for each IMF is extracted by spectrum analysis. A spectrum amplitude index is introduced as a method to select the optimal result. A comparison study of the method proposed here and some well-established techniques for detecting machinery faults is conducted through the analysis of both gear and bearing vibration signals. The results indicate that the proposed method has superior capability to extract machine fault features from vibration signals.

[1]  Yuesheng Xu,et al.  A B-spline approach for empirical mode decompositions , 2006, Adv. Comput. Math..

[2]  B. D. Forrester.  Gear Fault Detection using the Wigner-Ville Distribution , 1990 .

[3]  Anand Parey,et al.  Dynamic modelling of spur gear pair and application of empirical mode decomposition-based statistical analysis for early detection of localized tooth defect , 2006 .

[4]  Yang Yu,et al.  A fault diagnosis approach for roller bearings based on EMD method and AR model , 2006 .

[5]  D. Menicucci,et al.  Deriving the respiratory sinus arrhythmia from the heartbeat time series using empirical mode decomposition , 2003, q-bio/0310002.

[6]  Jean Claude Nunes,et al.  Image analysis by bidimensional empirical mode decomposition , 2003, Image Vis. Comput..

[7]  M. Zuo,et al.  Gearbox fault detection using Hilbert and wavelet packet transform , 2006 .

[8]  Robert B. Randall,et al.  THE RELATIONSHIP BETWEEN SPECTRAL CORRELATION AND ENVELOPE ANALYSIS IN THE DIAGNOSTICS OF BEARING FAULTS AND OTHER CYCLOSTATIONARY MACHINE SIGNALS , 2001 .

[9]  Yuesheng Xu,et al.  Gearbox fault diagnosis using empirical mode decomposition and Hilbert spectrum , 2006 .

[10]  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.

[11]  Isa Yesilyurt,et al.  Fault detection and location in gears by the smoothed instantaneous power spectrum distribution , 2003 .

[12]  Ming J. Zuo,et al.  GEARBOX FAULT DIAGNOSIS USING ADAPTIVE WAVELET FILTER , 2003 .

[13]  S. Loutridis.  Instantaneous energy density as a feature for gear fault detection , 2006 .

[14]  Cécile Capdessus,et al.  CYCLOSTATIONARY PROCESSES: APPLICATION IN GEAR FAULTS EARLY DIAGNOSIS , 2000 .

[15]  Shunian Yang,et al.  Improvement of the EMD method and applications in defect diagnosis of ball bearings , 2006 .

[16]  Wenyi Wang,et al.  Autoregressive Model-Based Gear Fault Diagnosis , 2002 .

[17]  Gabriel Rilling,et al.  On empirical mode decomposition and its algorithms , 2003 .

[18]  P. D. McFadden,et al.  APPLICATION OF WAVELETS TO GEARBOX VIBRATION SIGNALS FOR FAULT DETECTION , 1996 .

[19]  Wieslaw J. Staszewski,et al.  LOCAL TOOTH FAULT DETECTION IN GEARBOXES USING A MOVING WINDOW PROCEDURE , 1997 .