Method for confirming optimal resonance frequency band based on period target

Abstract

The invention discloses a method for confirming an optimal resonance frequency band based on a period target, and the method is characterized by adsorbing a vibration acceleration sensor to a tested rolling bearing pedestal to acquire vibration signals, creating a 1 / 3 binary tree filter bank to gain filtered complex envelope signals, calculating the periodical intensity PAR of each envelope signal, clearing the complex envelope signal kurtosis value less than a threshold value, calculating the kurtosis values of the surplus envelop signals, selecting the frequency band corresponding to the greatest-kurtosis complex envelope signal as the optimal resonance frequency band and the frequency spectrum of the envelope signal so as to obtain an envelop spectrum, in comparison with the fault feature frequency corresponding to the fault type in the rolling bearing, finally confirming the fault type of the rolling bearing; the method utilizes the substantive characteristics of the fault stimulation impact to avoid the non-periodic impact to interfere the resonance frequency demodulation band, and the method has robustness and can accommodate the confirmation process of the resonance frequency demodulation band so that fault feature extraction and diagnostic test can be automatically achieved.

Description

Technical field [0001] The invention relates to the field of rolling bearing fault diagnosis technology and signal processing analysis technology, in particular to a method for determining an optimal resonance frequency band based on period targeting. Background technique [0002] Resonance demodulation method is one of the most commonly used vibration and shock feature extraction technology, which is widely used in the monitoring and diagnosis of rolling bearings of large and key mechanical equipment. Resonance demodulation method first uses percussion test or spectral kurtosis method to determine the resonance frequency band of the impact signal; on this basis, bandpass filtering technology is used to suppress the impact of test noise and low-frequency vibration components, and accordingly extract the impact component of rolling bearing Time-domain waveform; finally, the envelope spectrum analysis of the filtered signal can obtain the characteristic frequency of the bearing fau...

AI Technical Summary

Problems solved by technology

[0004] However, Australia Randall pointed out that the method of fast spectral kurtosis map is sensitive to large random shock disturbances

This kind of interference will cause the original method to select the wrong filter frequency band, and cannot achieve the purpose of resonance demodulation technology for fault feature extraction and diagnosis

In addition, the fault characteristics of rolling bearings are periodic shocks, and the fast spectral kurtosis map method only considers the shock (kurtosis value) of fault-induced shocks in the process of determining the filtering frequency band, while ignoring the periodicity

The above defects will not only cause the determined filter frequency band to be dominated by one or several large random impact interferences, but also fail to extract fault feature information for effective fault diagnosis

[0005] Patent application number 201410140890.6 aims at the problem of inaccurate selection of resonance demodulation frequency band caused by high-peak pulse interference, and proposes to optimize the selection of resonance frequency band by sub-band spectrum kurtosis averaging, but the method does not consider the impact signal periodicity, and the periodicity of the shock is the premise of the envelope demodulation of the resonant frequency band

Images