. In this study, a new method for bearing fault diagnosis using local characteristic-scale decomposition multi-scale permutation entropy (LCD-MPE) and extreme learning machine AdaBoost (ELM-AdaBoost) algorithms is proposed. Vibration signals of railway axle box rolling bearings under 4 conditions (normal, outer race fault, inner race fault, and rolling element fault) were used as our research objects. The signals were de-noised using wavelet de-noising (WD) as a pre-filter, then the LCD was used to decompose the signal into a number of intrinsic scale components (ISCs). Then, the multi-scale permutation entropy (MPE) was extracted as the feature parameters. Finally, the extracted features were used as ELM-AdaBoost to achieve the automated fault diagnosis. Our results prove that our method is effective for an accurate diagnosis of railway axle box bearing faults. Furthermore, our fault diagnosis method is highly applicable in practical engineering.
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|