Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2026
|
| Online Access: | https://doi.org/10.5281/zenodo.19674720 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866901748456620032 |
|---|---|
| author | Basem Abd-Elhamed Rashad Shazly Nasser Fahmy Wael Abdelfattah |
| author_facet | Basem Abd-Elhamed Rashad Shazly Nasser Fahmy Wael Abdelfattah |
| contents | <p><span class="fontstyle0">High-impedance faults (Hi-ZFs) in low-voltage distribution networks remain challenging to detect due to their nonlinear<br>characteristics and low fault currents. This study presents an efficient and robust framework for fault detection and classification,<br>integrating Fast Wavelet Transform–based signal analysis with a Radial Basis Function Neural Network (RBFNN) for intelligent<br>decision-making. The FWT is employed to extract discriminative time–frequency features from phase current signals, enabling effective<br>characterization of transient and steady-state fault signatures. These features are then processed by an RBFNN classifier, selected for<br>its rapid convergence and strong generalization capability, making it suitable for real-time protection applications. The proposed method<br>is evaluated on the unbalanced IEEE 13-Bus low-voltage distribution benchmark tested system, employing ATP/EMTP and<br>MATLAB/Simulink platforms under various operating conditions, including conventional faults (LIFs), high-impedance faults (Hi-ZFs),<br>load/capacitor switching events, and different fault locations and inception angles. Simulation results demonstrate high detection<br>accuracy, achieving a classification rate of </span><span class="fontstyle2">99.73</span><span class="fontstyle0">%, robust classification performance, and fast response time within </span><span class="fontstyle2">21.5 ms</span><span class="fontstyle0">,<br>outperforming conventional signal-processing-based and machine-learning-based approaches. The findings confirm the effectiveness of<br>the proposed FWT–RBFNN approach as a practical and computationally efficient solution for enhanced fault protection in modern<br>distribution networks.</span> </p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_19674720 |
| institution | Zenodo |
| language | |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | A Fast Wavelet–Radial Basis Function Neural Network Approach for High-Impedance Fault (Hi-ZFs) Detection and Classification in LowVoltage Distribution Networks Basem Abd-Elhamed Rashad Shazly Nasser Fahmy Wael Abdelfattah <p><span class="fontstyle0">High-impedance faults (Hi-ZFs) in low-voltage distribution networks remain challenging to detect due to their nonlinear<br>characteristics and low fault currents. This study presents an efficient and robust framework for fault detection and classification,<br>integrating Fast Wavelet Transform–based signal analysis with a Radial Basis Function Neural Network (RBFNN) for intelligent<br>decision-making. The FWT is employed to extract discriminative time–frequency features from phase current signals, enabling effective<br>characterization of transient and steady-state fault signatures. These features are then processed by an RBFNN classifier, selected for<br>its rapid convergence and strong generalization capability, making it suitable for real-time protection applications. The proposed method<br>is evaluated on the unbalanced IEEE 13-Bus low-voltage distribution benchmark tested system, employing ATP/EMTP and<br>MATLAB/Simulink platforms under various operating conditions, including conventional faults (LIFs), high-impedance faults (Hi-ZFs),<br>load/capacitor switching events, and different fault locations and inception angles. Simulation results demonstrate high detection<br>accuracy, achieving a classification rate of </span><span class="fontstyle2">99.73</span><span class="fontstyle0">%, robust classification performance, and fast response time within </span><span class="fontstyle2">21.5 ms</span><span class="fontstyle0">,<br>outperforming conventional signal-processing-based and machine-learning-based approaches. The findings confirm the effectiveness of<br>the proposed FWT–RBFNN approach as a practical and computationally efficient solution for enhanced fault protection in modern<br>distribution networks.</span> </p> |
| title | A Fast Wavelet–Radial Basis Function Neural Network Approach for High-Impedance Fault (Hi-ZFs) Detection and Classification in LowVoltage Distribution Networks |
| url | https://doi.org/10.5281/zenodo.19674720 |