Lkurtogram Guided Adaptive Empirical Wavelet Transform and Purified Instantaneous Energy Operation for Fault Diagnosis of Wind Turbine Bearing

Abstract

The periodic impacts are regarded as the typical characteristics of local defect of wind turbine bearing. For this reason, it is significant to extract the periodic impacts from the original measured vibration signal with background noise interferences during the defect identification process. For the purpose of effectively solving this issue, an innovative diagnostic frame based on Lkurtogram guided adaptive empirical wavelet transform (LGAEWT) and purified instantaneous energy operation (PIEO) is put forward. Within this diagnostic frame, the L-kurtosis indicator guided wavelet equivalent filter band mergence is carried out to confirm the optimal filter boundary, then the resonance frequency band containing rich characteristic information can be extracted, and the characteristic sensitive component is able to be separated from the original vibration signal. In addition, a novel signal processing strategy called PIEO is presented to inhibit the background interferences and enhance the periodic impact signatures in sensitive components, and the purified instantaneous energy spectrum is adopted to replace the traditional envelope spectrum for characteristic frequency spectral lines identification. The feasibility of the proposed diagnostic frame has been demonstrated by the experimental signals and the actual engineering case, and the results manifest that it is suitable for bearing fault enhance detection under low signal-to-noise ratio (SNR) environment. Furthermore, the comparison results with the widely used Kurtogram, Autogram, and DWT methods indicate that this innovative diagnostic strategy has more prominent advantages on background interference suppression and weak characteristic intensification.