An Auto Instantaneous Frequency Order Extraction Method for Bearing Fault Diagnosis under Time-Varying Speed Operation

Abstract

Bearing fault diagnosis under variable speed usually have confronted two obstacles: a) blurry time frequency representation (TFR) and thus unavailable instantaneous frequency (IF) for resampling, and b) errorprone resampling process. To address such problems, this paper proposes a method which consists of two main steps: a) a regional peak search algorithm which searches the frequency bins point by point at local frequency regions is developed to extract the IF from the TFR of the original signal, and b) with the accurate IF estimator (either shaft IF, instantaneous fault characteristic frequency (IFCF) or their harmonics), an order peak highlighting strategy is exploited via multi-demodulating the signal and superposing the resulted frequency spectra of all demodulated signal components which are acquired by adaptive band-pass filtering. Then the instantaneous frequency order (IFO) of signal components of interest contained in the original signal can be highlighted and the IFO spectrum can be obtained for bearing fault diagnosis. In this manner, the bearing fault can be diagnosed without the tachometer, predenosing and resampling involved, indicating that the proposed can substantially reduce human involvement and facilitate its implementation in a fault detection expert system. The effectiveness of the proposed method are validated by both simulated and experimental data.