No abstract

In low voltage distribution networks, there is a high risk of fires caused by wiring faults that lead to short circuits. Arcing protection devices (ARCs) already exist, but experiments show that they also often do not disable such faults. The operation of most existing arc fault detection devices is based on the detection of high frequency current components and their comparison with load profiles. For the fast arc fault recognition low-frequency oscillations can be assessed, which serve as a marker of the arc fault presence in a LV distribution network. This method of arc fault recognition in LV networks is easier to implement practically. Using the windowed Fourier transform or Wavelet analysis, it is possible to determine the change in time of the harmonic components of the current, which will be a characteristic sign of the presence of an arc fault along with other signs.

No abstract