Feature Extraction Method of Transformer Vibration Based on Ensemble Empirical Mode Decomposition Subband

To extract the abnormal features of transformer in the early stage, the envelope demodulation method based on ensemble empirical mode decomposition (EEMD) and mutual information criterion was proposed. Firstly, the transformer vibration signal is decomposed into several intrinsic mode functions (IMFs) by the EEMD method. Then the singular value decomposition (SVD) is used to estimate the number of vibration source and mutual information criterion is performed to select the IMFs. Finally, extract the vibration features of the transformer through envelope demodulation of the IMFs. The method has good adaptability and can extract the weak vibration signal. And the validity of the method is verified by analyzing the measured transformer vibration signal.