Gear Fault Diagnosis Method Using Ensemble Empirical Mode Decomposition Energy Distribution and Grey Similar Incidence

Abstract

Focusing on some correspondence between energy distribution of different frequency bands and fault types when gear fault occurs, a new comprehensive fault diagnosis method is proposed based on ensemble empirical mode decomposition and grey similar incidence. The idea of circle statistics is introduced to improve the shortcoming of traditional morphological filter; and the rank-order morphological filter is defined. The line structure element is selected for rank-order morphological filter to denoise the original acceleration vibration signal. Denoised vibration signals are decomposed into a finite number of stationary intrinsic mode functions and some containing the most dominant fault information are calculated the energy distribution. Due to the grey similar incidence has good classify capacity for small sample pattern identification; these energy distributions could serve as the feature vectors, the grey incidence of different gear vibration signals is calculated to identify the fault pattern and condition. Practical results show that the proposed method can be used in gear fault diagnosis effectively.