Arc fault detection for AC SSPC in MEA with HHT and ANN

The detection of arc faults for AC Solid State Power Controller (SSPC) in more electric aircraft (MEA) still remains a challenge, since it has to be done while SSPC is still in operation and such arc faults will not provide considerable fault features. In this paper, a method based on Hilbert-Huang transform (HHT) and artificial neural networks (ANN) is proposed for AC SSPC arc fault detection. The adopted method using empirical mode decomposition (EMD) to decompose complex arc transient signal into finite intrinsic mode signal (IMF), so that the instantaneous frequency of Hilbert-Huang transform will have real physical meaning, and then the extracted instantaneous amplitude of the IMF is selected as a feature vector of arc current. Specifically, Hilbert-Huang transform based multi-resolution analysis is adopted to obtain the features of the AC SSPC arc current in the measured signal, artificial neural networks is adopted to identify the faults based on the extracted features. Numerical simulation results together with discussions have also been provided which indicates the effectiveness of the proposed fault detection method.