EMD- WVD Method Based High- Frequency Current Analysis of Low Voltage Arc

Experiment shows that high-frequency pulse current exists when arc fault occurs, which is one of signal characteristics that can be used for arc fault detection or early warning of potential low voltage arc. Motivated by this, it's critical to make a detailed time-frequency analysis of the arc fault signal. In this paper, we present an empirical mode decomposition (EMD) and Wigner- Ville distribution (WVD) method based high-frequency current analysis of low voltage arc. To eliminate the cross-terms caused by only WVD, we firstly decompose the denoised signal into single frequency component by using EMD, then use WVD to analyze each single component. The final waveform can help study the relationship between attenuation of high-frequency current and distance or frequency. Experimental results demonstrate that the proposed method shows higher accuracy and reliability in signal characteristics analysis, laying foundation to further detection and localization of arc fault.