Time series forecasting based on optimized LLM for fault prediction in distribution power grid insulators

TL;DR

This paper introduces a hybrid deep learning model utilizing an optimized large language model for accurate short- and medium-term time series forecasting of leakage current in power grid insulators, aiding fault prediction.

Contribution

It presents a novel hybrid deep learning approach combining signal filtering, multi-criteria optimization, and an optimized large language model for fault prediction in electrical insulators.

Findings

Outperforms existing deep learning models in leakage current prediction

Achieves low root-mean-square error of 2.24×10^{-4} for short-term forecasts

Effective in medium-term fault prediction with an error of 1.21×10^{-3}

Abstract

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$\times10^{-4}$ for a short-term horizon and 1.21$\times10^{-3}$ for a medium-term horizon.