Short-term electricity load forecasting with multi-frequency reconstruction diffusion
Qi Dong, Rubing Huang, Ling Zhou, Dave Towey, Jinyu Tian, Jianzhou Wang
TL;DR
This paper addresses short-term electricity load forecasting (STELF) by introducing MFRD, a diffusion-model framework that fuses multi-frequency reconstruction with a denoising network. It decomposes load signals using Variational Mode Decomposition (VMD), combines the original data with frequency components, and applies forward diffusion followed by a Transformer-augmented, residual-LSTM denoiser to generate forecasts. The approach consistently outperforms strong baselines on AEMO and ISO-NE datasets, with ablation studies highlighting the contributions from multi-frequency reconstruction, the LSTM component, and Fourier-based loss. The work advances STELF by leveraging diffusion modeling to enhance noise robustness and captures multi-scale temporal dynamics, offering a robust method for practical deployment in dynamic grids, while also signaling opportunities to reduce computational cost and expand spatial modeling in future work.
Abstract
Diffusion models have emerged as a powerful method in various applications. However, their application to Short-Term Electricity Load Forecasting (STELF) -- a typical scenario in energy systems -- remains largely unexplored. Considering the nonlinear and fluctuating characteristics of the load data, effectively utilizing the powerful modeling capabilities of diffusion models to enhance STELF accuracy remains a challenge. This paper proposes a novel diffusion model with multi-frequency reconstruction for STELF, referred to as the Multi-Frequency-Reconstruction-based Diffusion (MFRD) model. The MFRD model achieves accurate load forecasting through four key steps: (1) The original data is combined with the decomposed multi-frequency modes to form a new data representation; (2) The diffusion model adds noise to the new data, effectively reducing and weakening the noise in the original data; (3) The reverse process adopts a denoising network that combines Long Short-Term Memory (LSTM) and Transformer to enhance noise removal; and (4) The inference process generates the final predictions based on the trained denoising network. To validate the effectiveness of the MFRD model, we conducted experiments on two data platforms: Australian Energy Market Operator (AEMO) and Independent System Operator of New England (ISO-NE). The experimental results show that our model consistently outperforms the compared models.