Enhancing grid reliability through Transformer-LSTM model in smart grids

In the evolving landscape of global energy and information technology sectors in the early 21st century, the concept of smart grids has emerged as pivotal for enhancing the efficiency and reliability of power systems. This paper explores the integration of modern communication, computer, network, and control technologies within smart grids, highlighting their role in ensuring grid reliability, intelligence, and responsiveness. Despite varying definitions globally, smart grids universally aim to optimize grid operations through advanced technology integration. This article introduces a novel Transformer-LSTM encoder-decoder structure that integrates LSTM 's robust capability to capture long-term dependencies with Transformer's proficiency in capturing global dependencies. The proposed model is applied to forecast fluctuations in power grid data traffic, facilitating real-time adjustments to the grid's operational and maintenance strategies. Experimental results validate that the traffic prediction accuracy of the Transformer-LSTM model exceeds that of the independent Transformer and LSTM models.