Feature Extraction of High Impedance Arcing Faults in Compensated MV Networks. Part II: DWT-Based Analysis of Residual Components

Electrical faults due to leaning trees have been studied in [1]-[2] and it was found that the initial transients due to the associated arc reignitions are behavioral traits. These transients have been localized using the discrete wavelet transform (DWT). In this paper, these features are extracted from residual voltages using DWT. Therefore, the fault localization is carried out. The term for locating the faulty feeder is based on the polarity of a specific frequency band-power computed by multiplying the DWT detail coefficient of the residual voltage and current for each feeder. The performance of the proposed technique is getting better when the residual waveforms are utilized. Test cases provide evidence of the efficacy of the proposed technique.