Leveraging LLMs for Energy Forecasting: The AcegasApsAmga Case Study

Kevin Roitero, Andrea Zancola, Vincenzo Della Mea, Stefano Mizzaro

Abstract: This paper presents AcegasApsAmga’s application of Large Language Models (LLMs) for energy forecasting, focusing on both short-term and long-term power consumption predictions. We detail the model adaptation process, including fine-tuning techniques specific to energy data, and the integration of temporal and contextual features using Retrieval Augmented Generation (RAG) to enhance forecasting accuracy.

External IDs:dblp:conf/ecir/RoiteroZMM25