High impedance fault detection in 11 kV overhead line with discrete wavelet transform and independent component analysis
| Autor: | Abdelrehman Omer Idris, Hussain Sharif, Rohaiza Hamdan, Ferdouse Hossain Bhuiya, Dur Mohammad Soomro |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
Discrete wavelet transform
Control and Optimization Computer Networks and Communications Computer science Noise (signal processing) Fault (power engineering) Independent component analysis Fault detection and isolation High impedance Moving window Hardware and Architecture Signal Processing Waveform HIF ICA Electrical and Electronic Engineering Fault model DWT Algorithm Information Systems |
| Popis: | This paper proposes an analysis of high-impedance fault detection algorithms for medium voltage distribution lines based on the discrete wavelet transform (DWT) technique and a more advanced technique named independent component analysis (ICA) independently. Three-phase distribution line model and two diodes high impedance fault model, which represents the unsymmetrical fault current of electric arc, simulated using MATLAB/Simulink. High impedance fault (HIF) detection algorithm initially analyzes the sampled current waveforms through DWT and the resultant third level high-frequency components “d3” coefficients are analyzed through one cycle moving window approach. The proposed algorithm successfully detects any HIF in the distribution current even if there is a slight or no difference in the amplitude of the HIF and the waveform of the phase current. On the other hand, the ICA more developed algorithm than DWT successfully separated the noise signals from the obtained current waveforms and HIF noise signals can be differentiated with non-HIF noise signals. Because of this reason ICA is chosen in this research. The detected HIF current can be from 50 ma and up. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|