A new rotor fault identification method based on normalized energy distribution and grey relation degree

In this paper, a new fault identification method was proposed based on normalized energy distribution and grey relation degree. Firstly, the sampled data was denoised by morphological filter. Secondly, the denoised signal was transformed in frequency domain, and some frequency bands which could express the typical fault type were selected by harmonic window decomposition. Thirdly, the normalized energy distributions of these frequency bands were calculated. Different fault type corresponds with different normalized energy distribution. Finally, the grey relation between the symptom set and standard fault set was calculated as the identification evidence for fault diagnosis. The practical results show that this method is quite effective in rotor fault identification. It´s suitable for on-line monitoring and diagnosis of rotating machinery.