Partial Discharge Pattern Recognition Using Hilbert – Huang Transform in Acoustic Signal Analysis

This paper proposes new partial discharge (PD) pattern recognition based on the Hilbert–Huang Transform (HHT). First, five types of defect models are well-designed on the base of investigation of many power equipment failures, and a commercial acoustic emission (AE) sensor measures the acoustic signal caused by the PD phenomenon. Next, the HHT can represent instantaneous frequency components through empirical mode decomposition (EMD), then transform to a 3D energy spectrum. Finally, the feature parameters of energy are extracted from the 3D Hilbert spectrum, using a neural network (NN) for PD recognition. To demonstrate the effectiveness of the proposed method, the identification ability is investigated on 150 sets of field-tested PD patterns of defect models. The proposed method can easily separate different defect types and shows good tolerance when random noise is added.