An arc fault diagnostic method for low voltage lines using the difference of wavelet coefficients

It is difficult to distinguish between fault arc and similar arc generated by the special load during normal working condition using the existing arc fault detection method. In this paper, a new arc fault detecting method of fast wavelet transform based on multi-resolution analysis is proposed. In the method, the sampling loads current was decomposed to low frequency signal and high frequency signals using the Mallat algorithm, and the mean and its difference of the reconstructed wavelet high frequency coefficients are treated as the criteria for arc fault diagnosis. Compared with the existing detection methods, this method can significantly reduce the computational burden and has high diagnostic reliability. The current characteristics of various loads are analyzed in normal or abnormal situations by the experiment to verify the accuracy of the method. The experimental results show that, the detection method can differentiate between fault arc and similar arc accurately, and the identification ability for specific loads is enhanced. The detection method can be widely used in low voltage power supply and distribution system to improve the diagnostic accuracy of the arc fault circuit interrupter.