Detecting of Transient Vibration Signatures Using an Improved Fast Spatial–spectral Ensemble Kurtosis Kurtogram and Its Applications to Mechanical Signature Analysis of Short Duration Data from Rotating Machinery
BinQiang Chen,ZhouSuo Zhang,YanYang Zi,ZhengJia He,Chuang Sun
DOI: https://doi.org/10.1016/j.ymssp.2013.03.021
IF: 8.4
2013-01-01
Mechanical Systems and Signal Processing
Abstract:Detecting transient vibration signatures is of vital importance for vibration-based condition monitoring and fault detection of the rotating machinery. However, raw mechanical signals collected by vibration sensors are generally mixtures of physical vibrations of the multiple mechanical components installed in the examined machinery. Fault-generated incipient vibration signatures masked by interfering contents are difficult to be identified. The fast kurtogram (FK) is a concise and smart gadget for characterizing these vibration features. The multi-rate filter-bank (MRFB) and the spectral kurtosis (SK) indicator of the FK are less powerful when strong interfering vibration contents exist, especially when the FK are applied to vibration signals of short duration. It is encountered that the impulsive interfering contents not authentically induced by mechanical faults complicate the optimal analyzing process and lead to incorrect choosing of the optimal analysis subband, therefore the original FK may leave out the essential fault signatures. To enhance the analyzing performance of FK for industrial applications, an improved version of fast kurtogram, named as “fast spatial–spectral ensemble kurtosis kurtogram”, is presented. In the proposed technique, discrete quasi-analytic wavelet tight frame (QAWTF) expansion methods are incorporated as the detection filters. The QAWTF, constructed based on dual tree complex wavelet transform, possesses better vibration transient signature extracting ability and enhanced time-frequency localizability compared with conventional wavelet packet transforms (WPTs). Moreover, in the constructed QAWTF, a non-dyadic ensemble wavelet subband generating strategy is put forward to produce extra wavelet subbands that are capable of identifying fault features located in transition-band of WPT. On the other hand, an enhanced signal impulsiveness evaluating indicator, named “spatial–spectral ensemble kurtosis” (SSEK), is put forward and utilized as the quantitative measure to select optimal analyzing parameters. The SSEK indicator is robuster in evaluating the impulsiveness intensity of vibration signals due to its better suppressing ability of Gaussian noise, harmonics and sporadic impulsive shocks. Numerical validations, an experimental test and two engineering applications were used to verify the effectiveness of the proposed technique. The analyzing results of the numerical validations, experimental tests and engineering applications demonstrate that the proposed technique possesses robuster transient vibration content detecting performance in comparison with the original FK and the WPT-based FK method, especially when they are applied to the processing of vibration signals of relative limited duration.