Advancing Generative Artificial Intelligence and Large Language Models for Demand Side Management with Internet of Electric Vehicles

원문 초록

Generative artificial intelligence, particularly through large language models (LLMs), is poised to transform energy optimization and demand side management (DSM) within microgrids. This paper explores the integration of LLMs into energy management, emphasizing their roles in automating the optimization of DSM strategies with Internet of electric vehicles. We investigate challenges and solutions associated with DSM and explore the new opportunities presented by leveraging LLMs. Then, we propose an innovative solution that enhances LLMs with retrieval-augmented generation for automatic problem formulation, code generation, and customizing optimization. We present a case study to demonstrate the effectiveness of our proposed solution in charging scheduling and optimization for electric vehicles, highlighting our solution's significant advancements in energy efficiency and user adaptability. This work underscores the potential of LLMs for energy optimization and fosters a new era of intelligent DSM solutions.