Predicting Grounding Resistance of Transmission Towers for Enhanced Lightning Protection Using Temporal Fusion Transformer

The safe and stable operation of power grids is essential to meet the growing demand for electricity in modern society. Transmission lines, crucial to the power system, are particularly vulnerable to damage from lightning strikes, which account for a significant portion of power grid failures, especially in mountainous regions with complex terrain. Grounding resistance plays a vital role in lightning protection, with its performance being influenced by various environmental and operational factors. This paper proposes a novel approach for predicting the grounding resistance of transmission towers using the Temporal Fusion Transformer (TFT) model. TFT, an advanced attention-based deep learning model, is employed to forecast grounding resistance values by leveraging temporal and environmental data. By integrating static covariates, observed inputs, and known future inputs, the TFT model captures both short-term and long-term dependencies, enabling accurate predictions of grounding resistance across different scenarios. Experimental results demonstrate the potential of the TFT model to enhance the reliability of lightning protection systems and contribute to the safe operation of transmission lines, with significant implications for power grid maintenance and disaster prevention strategies.