A Rolling Bearing Fault Diagnosis Technique Based on WOA-VMD-SVM

Variational Mode Decomposition (VMD) requires the manual setting of decomposition algorithm parameters, leading to low fault identification accuracy. To address this issue, this paper proposes a fault diagnosis method based on the Whale Optimisation Algorithm (WOA)-optimized VMD, combined with Multi-scale Permutation Entropy (MPE) and Support Vector Machine (SVM). The specific steps are as follows: Firstly, the Whale Optimisation Algorithm is introduced to optimize the number of modal decompositions (K) and the quadratic penalty factor (alpha) of Variational Mode Decomposition, obtaining the optimal VMD parameters. After decomposing the vibration signal, the multi-scale permutation entropy of each Intrinsic Mode Function (IMF) component is calculated separately. Finally, the extracted feature data are input into the SVM classifier, and the Whale Optimisation Algorithm is used to optimize the parameters of SVM to obtain the WOA-SVM optimization model. Meanwhile, ten-fold cross-validation is introduced to train the WOA-SVM, ultimately achieving accurate classification and diagnosis of fault signals. Analyses are conducted using datasets from the Bearing Data Centers of Jiangnan University and the University of Western Ontario (Canada) respectively. The results show that the fault identification accuracy of this method reaches 98.36% and 99.13%, respectively, and its performance is superior to that of methods such as VMD-PE, EMD-MPE, VMD-SVM, and VMD-MPE.