Mechanical Fault Diagnosis of Transformer Windings Based on VMD and DBO-SVM
Quan Fa Zhou, Ziqing Ye, Zhong Sen, Yuru Pan · 2024
In order to diagnose mechanical faults of transformer winding looseness more accurately and effectively, a transformer winding looseness fault diagnosis method based on Variational Mode Decomposition (VMD) and Dung Beetle Optimization Algorithm optimized Support Vector Machine (DBO-SVM) is proposed. Firstly, conduct a simulated fault experiment on a 10 kV transformer and measure its vibration signal. Subsequently, the non-stationary vibration signal is decomposed into multiple intrinsic mode functions (IMFs) using VMD, and the energy entropy of each layer of IMF is calculated to form an eigenvector. Finally, the feature vectors are input into the SVM optimized by the DBO to train a classification model and achieve transformer winding loosening fault diagnosis. The research results indicate that the method proposed in this article is suitable for diagnosing loose faults in transformer windings, and its fault recognition accuracy is higher compared to the traditional improved SVM classification model.