A novel wavelet packet transform based fault identification procedures in HV transmission line based on current signals

Abstract

This paper presents an adaptive fault identification algorithm bases on wavelet packet transform (WPT) for two-terminal power transmission lines. The proposed scheme performs four functions which are the fault detection, fault classification, distinguishing among the temporary and the permanent faults, and detection of the arc extinguish instant. The presented algorithm only uses the measured current at one terminal reducing the required cost. Also, it can mitigate the error resulting from the load variations via updating the presetting value. Consequently, it does not need retesting under changing the transmission system configurations. The proposed scheme is deduced in the spectral domain and depended on the application of the WPT. The db6 wavelet packet is used for decomposing the faulty phase current waveform (level 7) to get the energy coefficients. The presented algorithm is assessed under various fault conditions such as fault distances, inception angles, and faults nature via simulating different secondary arc models via using ATP/EMTP. The obtained results are investigated and evaluated.