Emergent Mind

Abstract

Accurate load forecasting is essential for maintaining the power balance between generators and consumers, especially with the increasing integration of renewable energy sources, which introduce significant intermittent volatility. With the development of data-driven methods, machine learning and deep learning-based models have become the predominant approach for load forecasting tasks. In recent years, pre-trained language models (PLMs) have made significant advancements, demonstrating superior performance in various fields. This paper proposes a load forecasting method based on PLMs, which offers not only accurate predictive ability but also general and flexible applicability. Additionally, a data modeling method is proposed to effectively transform load sequence data into natural language for PLM training. Furthermore, we introduce a data enhancement strategy that eliminate the impact of PLM hallucinations on forecasting results. The effectiveness of the proposed method has been validated on two real-world datasets. Compared with existing methods, our approach shows state-of-the-art performance across all validation metrics.

We're not able to analyze this paper right now due to high demand.

Please check back later (sorry!).

Generate a summary of this paper on our Pro plan:

We ran into a problem analyzing this paper.

Newsletter

Get summaries of trending comp sci papers delivered straight to your inbox:

Unsubscribe anytime.