| Autor: |
Miguel Jimenez-Aparicio, Felipe Wilches-Bernal, Matthew J. Reno |
| Jazyk: |
angličtina |
| Rok vydání: |
2023 |
| Předmět: |
|
| Zdroj: |
IEEE Access, Vol 11, Pp 74201-74215 (2023) |
| Druh dokumentu: |
article |
| ISSN: |
2169-3536 |
| DOI: |
10.1109/ACCESS.2023.3296737 |
| Popis: |
This work proposes a signal-based fault location method combining both time and frequency-domain features. The method is based on the Traveling Wave (TW) detection and analysis, and only employs 100 microseconds of data around the TW time-of-arrival to the measuring location. Time-domain features are extracted using Mathematical Morphology, while frequency-domain features are obtained using the Stationary Wavelet Transform. The IEEE 34 nodes system is selected for demonstration. TW detection is performed using a method based on Dynamic Mode Decomposition. This work exhaustively analyzes the performance of the proposed method in both fault location classification and regression tasks. Combining both time and frequency-domain features is proven to be more effective than using each type of feature independently. In addition, the sensitivity to noisy measurements and lesser amount of training data are analyzed as well. The performance of this method is compared to other works in the same test system, showing better accuracy than other signal-based methods. Furthermore, the proposed method has a low fault location estimation error rate across the entire length of the feeder and shows a similar performance to slower phasor-based fault location methods. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
