Extraction of Partial Discharge Signal Feature Based on Dual-tree Complex Wavelet Transform and Singular-value Decomposition

Different high-voltage cable fault can cause multiple high-frequency signal characterization. Effectively getting features of the signal can recognize Partial Discharge (PD) with the effect of de-noising and dimension reduction. In this paper, we first proceed to analyze signal based on dual-tree complex wavelet transform. Then we extract singularity signal by Singular value decomposition (SVD). To accomplish getting and recognizing features of PD, we set maximum singular value, average singular value and information entropy as characteristic quantity to analyze its distribution. The simulation results show that this method is feasible.