Fault Diagnosis for Metal Particles within GIL Based on VMD-SVM

Aiming at the problem of the low accuracy of classification and identification for metal particle faults in GIL, a metal particle fault diagnosis scheme is proposed in this study. First, Variational mode decomposition (VMD) is used to decompose the mode of vibration signals of metal particle fault. Second, the Hurst exponent of each intrinsic mode function (IMF-H) is computed by the R/S method. Third, input the IMF-H as the feature vector input into the SVM diagnostic model to classify and identify the metal particle faults. The results indicate that the method proposed in this study can effectively extract the fault features and improve the classification accuracy of fault diagnosis.