Abstract
Rolling element bearings are crucial components in all kinds of rotating machinery. Its fault detection is of great importance, as it ensures the performance of the whole machine. Periodic transient impulses caused by bearing defects are usually submerged in strong background noise which poses a challenge for effective fault feature extraction. To detect bearing faults reliably, a new fault feature extraction method is presented. First, the adaptive maximum second-order cyclostationary blind deconvolution is utilized to recover bearing fault-related impulses, while the optimal filter length is chosen based on the harmonic significance index which quantifies the diagnostic information contained in a deconvoluted signal. Second, cross-correlation is calculated between the teager energy operator and the envelope of the deconvoluted signal to further eliminate the irrelevant noise. Finally, fast fourier transform is employed to acquire the cross-correlation spectrum and the fault features can be extracted successfully. The performance of the proposed method is verified on both simulation signals and experimental signals acquired from a test rig. The superior abilities of noise reduction and fault detection are shown clearly when compared with some traditional method.
Highlights
Rolling element bearings (REBs) are widely used in rotating machinery and play an important role in modern industries
2.2 Adaptive filter length selection method based on harmonic significant index As the analyzed result of CYCBD for a vibration signal with bearing defect is influenced by the filter length and the longer filter length may lead to more time during convergence, the proper determine of filter length is the key point in the blind deconvolution (BD) process
To further demonstrate the validity of the proposed method, first, two other blind deconvolution approaches named minimum entropy deconvolution (MED) and maximum correlated kurtosis deconvolution (MCKD) are used. For both the two comparison methods, the filter length is selected equal to the optimal length of the CYCBD, and for the MCKD, the period of deconvolution is determined according to the fault characteristic frequency and sample frequency, the number of shift is chosen to be M=3
Summary
Rolling element bearings (REBs) are widely used in rotating machinery and play an important role in modern industries. The MCKD entails rigorous requires for input parameters, only when all the parameters are set properly, can the superiority of MCKD be highlighted To overcome this limitation, the cuckoo search algorithm (CSA) that aims to achieve the maximum feature energy ratio is employed to select the optimal filter length and number of shift, which was proved to be effective compared with some traditional methods [23]. To extract the weak fault features from the vibration signals masked by strong noise, a new method named ACYCBD-based cross-correlation spectrum is proposed. The main contribution of this paper is the use of harmonics significance index (HSI) to choose the optimal filter length for the CYCBD and the propose of cross-correlation spectrum to further enhance fault features of the deconvoluted signals.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
