PeakFocus: Bridging Peak Localization and Intensity Regression via a Unified Multi-Scale Framework for Electricity Load Forecasting

2026-05-20

AI Summary

This work addresses the limitations of existing two-stage approaches to electrical load peak forecasting, which struggle to jointly optimize peak timing localization and intensity regression due to multi-scale representation conflicts and temporal misalignment. To overcome these challenges, the authors propose PeakFocus, a unified end-to-end peak-aware framework that integrates a multi-scale hybrid peak locator with a position-aware decoder. By leveraging cascaded feature fusion and positional context injection, PeakFocus enables synergistic optimization of localization and regression tasks. The method is further enhanced by a triple-hybrid loss function and a tolerance-aware evaluation protocol. Experimental results on the ELC and WLEL datasets demonstrate that PeakFocus significantly outperforms current baselines, achieving state-of-the-art performance in both peak timing accuracy and intensity estimation.

Abstract

Electricity load peak forecasting (ELPF), simultaneously predicting peak timing and intensity, is a prerequisite for effective grid scheduling and risk management. However, existing methods face three limitations. First, they adopt a two-stage predict-then-locate paradigm, which severs the link between temporal localization and intensity regression. Second, they still struggle with the multi-scale representation conflict, leading to peak misjudgment and timing misalignment. Third, the lack of explicit peak timing context during intensity regression causes intensity smoothing because predictions are dominated by global smoothing trends. To address these limitations, we propose PeakFocus, a unified framework for ELPF. (i) A Unified Peak-Aware Pipeline (UPAP) utilizes a triple hybrid loss to jointly supervise temporal localization and intensity regression, alongside a tolerance-based evaluation protocol. (ii) A Multi-Scale Mixing Peak Locator (MSM-PL) exploits coarse-grained features to mitigate peak misjudgment caused by local fluctuations, and injects them into fine-grained features via a cascade mechanism to resolve timing misalignment. (iii) A Location-Aware Decoder (LAD) injects peak timing context into the intensity regression process, providing explicit guidance to counteract intensity smoothing and improve peak intensity estimation. Extensive experiments on the public Electricity (ELC) dataset and our industrial-scale World Large-scale Electricity Load (WLEL) dataset show that PeakFocus outperforms baselines in both timing precision and intensity estimation.

Problem

• Electricity load peak forecasting
• Peak localization
• Intensity regression
• Multi-scale representation
• Temporal alignment

Innovation

• peak-aware forecasting
• multi-scale representation
• unified framework
• temporal localization
• intensity regression