Series Arc Fault Detection Using Discrete Wavelet Transform

Arc faults mainly occur because of electrical problems such as aging cables and loose connections. Generating high temperature and discharging molten metal, arc faults finally lead to electrical fires. Every year such fires bring great loss and damage. It is identified that conventional protecting technique is unable to break a circuit in the situation when a brief arc fault occurs and the arc current is below thermal or instantaneous trip levels. In this paper a new method is proposed for the detection of residential series arc fault. Firstly, suitable arc model is used to simulate series arc fault. Then based on fault detection algorithm the fault features are extracted using a signal processing technique called Discrete Wavelet Transform (DWT). Then by comparing different wavelets, db4 were found appropriate for extracting arc-fault features. The last feature also involved the use of thresholding, peak detection, and relay triping. MATLAB were used to build and simulate arc-fault model.