Exportar Publicação

Matos-Carvalho, João P., Stefenon, Stefano Frizzo, Valderi Leithardt & Kin-Choong Yow (2025). Time series forecasting based on optimized LLM for fault prediction in distribution power grid insulators. arXiv is a free distribution service and an open-access archive for nearly 2.4 million scholarly articles,. https://doi.org/10.48550/arXiv.2502.17341

M. J. P. et al.,  "Time series forecasting based on optimized LLM for fault prediction in distribution power grid insulators", in arXiv is a free distribution service and an open-access archive for nearly 2.4 million scholarly articles, 2025

@misc{p.2025_1790104489225,
	author = "Matos-Carvalho, João P. and Stefenon, Stefano Frizzo and Valderi Leithardt and Kin-Choong Yow",
	title = "Time series forecasting based on optimized LLM for fault prediction in distribution power grid insulators",
	year = "2025",
	doi = "10.48550/arXiv.2502.17341",
	howpublished = "Digital",
	url = "https://arxiv.org/abs/2502.17341"
}

TY  - GEN
TI  - Time series forecasting based on optimized LLM for fault prediction in distribution power grid insulators
T2  - arXiv is a free distribution service and an open-access archive for nearly 2.4 million scholarly articles
AU  - Matos-Carvalho, João P.
AU  - Stefenon, Stefano Frizzo
AU  - Valderi Leithardt
AU  - Kin-Choong Yow
PY  - 2025
DO  - 10.48550/arXiv.2502.17341
UR  - https://arxiv.org/abs/2502.17341
AB  - Surface contamination on electrical grid insulators leads to an increase in leakage current until an electrical discharge occurs, which can result in a power system shutdown. To mitigate the possibility of disruptive faults resulting in a power outage, monitoring contamination and leakage current can help predict the progression of faults. Given this need, this paper proposes a hybrid deep learning (DL) model for predicting the increase in leakage current in high-voltage insulators. The hybrid structure considers a multi-criteria optimization using tree-structured Parzen estimation, an input stage filter for signal noise attenuation combined with a large language model (LLM) applied for time series forecasting. The proposed optimized LLM outperforms state-of-the-art DL models with a root-mean-square error equal to 2.24×10−4 for a short-term horizon and 1.21×10−3 for a medium-term horizon.
ER  -