New bearing slight degradation detection approach based on the periodicity intensity factor and signal processing methods

Abstract

In this article, a new approach has been presented in which novel hybrid features are introduced to detect the time of the beginning of the bearing degradation in the run-to-failure test. At first, the ensemble empirical mode decomposition (EEMD) method and the wavelet packet decomposition (WPD) are used in signal decomposition. The conventional frequency and time domain features, the envelope harmonic-to-noise ratio (EHNR) and the recently introduced periodicity intensity factor (PIF) are utilized for constructing the features matrix. Consequently, the most sensitive feature is identified using the compensation distance evaluation technique. The results show that the hybrid features obtained by the PIF, the EEMD and WPD methods can determine the exact moment of degradation. Also, the new hybrid features are superior to the features introduced in other studies in the early fault detection.