Abstract: The correlation dimension and wavelet energy spectrum entropy characteristics of rubbing fault signals were extracted, and the rubbing fault diagnosis model was built based on SVM. The variation rule of the correlation dimension and wavelet energy spectrum entropy of signals with rubbing faults was analysed by using a rubbing dynamic simulation model The simulated samples as the study samples were taken to train the SVM classifier and then the SVM discriminant function was obtained which can be used to identify the rubbing faults Finally, do experiments with the aero-engine rotor failure test rig to get rub-impact fault samples, and tjeir correlation dimension and energy entropy were calculated samples which are then substituted into the SVM discriminant function to determine rubbing fault directly. The experimental results verify the effectiveness of the method.
王美令;陈 果. 基于关联维数和小波能量谱熵的碰摩故障智能诊断[J]. , 2010, 29(8): 174-177,.
WANG Meiling;CHEN Guo. Intelligent Diagnosis of Rubbing Based on Correlation Dimension and Energy Spectrum Entropy. , 2010, 29(8): 174-177,.