Rotor Crack Detection Based on Multi-Vibration Signal Fusion Collected from the Basement of Machinery Using SVM and Statistical Characteristics Methodology
As the poor generability of special sensor support frame and the inconvenience of signal acquisition in the process of common fault diagnosis for cracked rotor, a new fault diagnosis method is presented in this paper. this method takes the basement of rotor test rig as the monitoring objects and makes feature fusion for time-domain statistics of multiple sensors using SVM (support vector machine). The result of experiment showed that the method using the multi-sensor signal fusion technology collected from the basement of machinery has the advantages of better diagnostic precision for rotor crack diagnosis, furthermore, it supplies a new way for rotor fault diagnosis.
Shengyi Li, Yingchun Liu, Rongbo Zhu, Hongguang Li, Wensi Ding
X. J. Li et al., "Rotor Crack Detection Based on Multi-Vibration Signal Fusion Collected from the Basement of Machinery Using SVM and Statistical Characteristics Methodology", Applied Mechanics and Materials, Vols. 34-35, pp. 1000-1004, 2010