원문정보
초록
영어
For rolling bearing fault, there is a certain mapping relationship between failure state and its each frequency band energy and distribution, this paper proposes a bearing fault diagnosis method based on wavelet packet decompotion energy distribution and gray similar correlation degree. Introducing a concept of soft morphological filtering, combined the EMD method on the original vibration signal de-noising pretreatment. Decomposing bearing vibration acceleration signal after noise reduction into different frequency band signal by adopting the method of wavelet packet decomposition and calculating the energy distribution. Due to the gray similar correlation degree analysis for the classification of the small sample pattern recognition has a good effect, element structure characteristic vector by energy distribution, to judge the fault types by calculating the bearing vibration signal under different status similar correlation degree. Experimental results show that, the proposed can be effectively applied to bearing fault diagnosis.
목차
1. Introduction
2. Preprocessing Principle of Flexible Morphological Filtering Noise based on EMD
3. Wavelet Packet Decomposition
4. Gray Similarity Degree
5. Fault Recognition of bearing based on Wavelet Packet Decomposition Energy Distribution and Gray Similarity Degree
6. Application
7. Conclusion
References