FAULT FEATURE EXTRACTION OF MECHANICAL EQUIPMENTS BASED ON ICA AND HHT
Vibration signals often contain plentiful running condition and fault information of mechanical equipment, but the noise and interference from environment or other equipments often disturb the feature extraction and fault diagnosis. If the time domain signals are effectively separated or decomposed, the useful components are extracted, and the interference from other components are restrained, the signal to noise ratio and the quality of diagnostic information will be improved. Aiming at large rotating machinery, independent component analysis (ICA) and Hilbert-Huang transformation (HHT) are applied to separate and decompose the vibration signals of mechanical equipment. Firstly, the mixture vibration signals are separated into different independent components according to different vibration source by ICA, and then decompose each independent component by EMD to several intrinsic mode functions. With the Hilbert transform, the IMFs yield instantaneous frequencies as functions of time that give sharp identifications of imbedded structures. The final presentation of the results is an energy-frequency-time distribution, designated as the Hilbert-Huang spectrum.The combined ICA-HHT method is applied in the monitoring and fault diagnosis of large rotate machine. The results show that this method can extract the fault information from mixed vibration signals, thereby providing an effective technology for condition monitoring and fault diagnosis of mechanical equipment.
Independent Component Analysis Hilbert-Huang Transformation Feature Extraction Fault Diagnosis
Xu YongGang Chang Hai Gao LiXin He JinQun
The Key Laboratory of Advanced Manufacturing Technology, College of Mechanical Engineering and Applied Electronics Technology, Beijing University of Technology, Beijing, 100124, China
国际会议
北京
英文
1803-1811
2008-10-27（万方平台首次上网日期，不代表论文的发表时间）