Real Time Series DC Arc Fault Detection Based on Fast Fourier Transform

Abstract

Now day’s the consumption of electricity increasing rapidly. One of them is a low voltage direct current (LVDC) system. Generally an LVDC installation system consists of several 100V DC power supply with current capacity under 10 A. The deadly dangers that arise in the system is arcing Fault. Series Arc Fault becomes the frequent Arc dangers caused by loose connections between cables and poor connections on installations. Different from over current indication, arcing fault are generally accompanied by the appearance of strange and unusual current waveforms tending with decreases current amplitude. Because of the uniqueness, the conventional current protection system cannot detect arcing fault. The solution is to apply time-frequency analysis to detect interference through the waveform of the current system. Intentionally a FFT method is used to decompose the waveform of the current system from time domain into frequency domain. The spectrum from FFT calculation will be used to recognize the current interference and determine arcing fault condition on the system. The experimental data shows that AFCI base on FFT method is able to detect series DC arc fault and give protection to the system with average of clearing time of 530 ms.