Artificial Neural Network-Based UHF Partial Discharge Localisation in Mineral Oil with Datasets Based on TDOA
Pertanika Journal of Science and Technology · 30.10.2026 · 34 (5)
DOI: 10.47836/pjst.34.5.03 ↗Kurz gefasst
Fachartikel in der Zeitschrift Pertanika Journal of Science and Technology, online erschienen am 30.10.2026. Verfasst von Yang Lim Zhi und 5 weiteren Autor:innen.
Abstract (Originalsprache)
This study presents a dataset generation procedure to train artificial neural networks (ANNs) for ultra-high frequency (UHF) partial discharge (PD) localisation in an oil-filled tank. The datasets were generated using the Time Difference of Arrival (TDOA) equations and augmented with samples containing random deviations. The PD localisation performance of the ANNs trained on the datasets was compared with that of the artificial bee colony (ABC), bat algorithm (BA), genetic algorithm (GA), particle swarm optimisation (PSO), and a non-iterative algorithm, based on the TDOAs of experimental UHF PD signals. An experiment was conducted in which UHF PD signals were captured with 4 patch-type wideband antennas and denoised using the variational mode decomposition (VMD) method. TDOAs were computed from both the original and denoised UHF PD signals through the First Peak Method (FPM) and Cumulative Energy Method (CEM). For the original UHF PD signals, the FPM produces TDOAs with lower mean absolute errors (MAEs) than the CEM. In contrast, the CEM produces TDOAs with lower MAEs than the FPM for the denoised UHF PD signals. TDOAs with lower MAEs generally lead to lower PD localisation errors. The trained ANNs consistently achieve the highest PD localisation accuracy, with average computation times ranging from 36.75 ms to 43.17 ms.
Zitieren
APA
Zhi, Y. L., Azis, N., Ramli, H. R., Radzi, M. A. M., Norsahperi, N. M. H., & Ariffin, A. M. (2026). Artificial Neural Network-Based UHF Partial Discharge Localisation in Mineral Oil with Datasets Based on TDOA. Pertanika Journal of Science and Technology, 34(5). https://doi.org/10.47836/pjst.34.5.03
Vancouver
Zhi YL, Azis N, Ramli HR, Radzi MAM, Norsahperi NMH, Ariffin AM. Artificial Neural Network-Based UHF Partial Discharge Localisation in Mineral Oil with Datasets Based on TDOA. Pertanika Journal of Science and Technology. 2026;34(5). doi:10.47836/pjst.34.5.03
BibTeX
@article{Zhi2026,
author = {Zhi, Yang Lim and Azis, Norhafiz and Ramli, Hafiz Rashidi and Radzi, Mohd Amran Mohd and Norsahperi, Nor Mohd Haziq and Ariffin, Azrul Mohd},
title = {Artificial Neural Network-Based UHF Partial Discharge Localisation in Mineral Oil with Datasets Based on TDOA},
journal = {Pertanika Journal of Science and Technology},
year = {2026},
volume = {34},
number = {5},
doi = {10.47836/pjst.34.5.03}
}
EUVIMED