IBAS-SVM Rolling Bearing Fault Diagnosis Method Based on Empirical Modal Characteristics

To solve the problem of incomplete feature extraction and difficulty in selecting optimal support vector machine parameters, a fault diagnosis method based on the ensemble empirical modal decomposition (EEMD) fuzzy entropy and improved beetle antennae search algorithm optimized SVM was proposed. After the original signal was decomposed by EEMD, several Intrinsic mode functions (IMF) were filtered by kurtosis and correlation coefficients to characterize the fault information, then the fuzzy entropy was extracted to reduce information redundancy and improve fault identification accuracy. At the same time, information sharing characteristic and adaptive step operator were introduced to improve the search capability of the beetle antennae search (BAS) algorithm. The experimental result shows that the average recognition accuracy of the method are 95.8% and 96.7% on CWRU and IMS datasets, which proves that our model can effectively extract the fault features and determine the rolling bearing fault types more accurately.