Uncertainty-aware probabilistic energy forecasting for low-carbon industrial energy management systems
低炭素産業エネルギー管理システムのための不確実性を考慮した確率的エネルギー予測 (AI 翻訳)
Divyabhan S. Duggal, Haimeng Wu, Neil S. Beattie, Anthony Browne
🤖 gxceed AI 要約
日本語
本論文は、産業用エネルギー管理システムにおける再生可能エネルギー導入時の不確実性に対応するため、多出力量子化LSTMアーキテクチャに基づく確率的予測フレームワークを提案する。コンフォーマル予測を用いて予測区間の信頼性を向上させ、PV発電、系統電力輸入、炭素強度の確率的予測を実現。実データによる評価では、冬期と夏期の両条件で堅牢な性能を示し、確率的ディープラーニングの産業エネルギー計画への有効性を確認した。
English
This paper proposes a probabilistic multi-horizon forecasting framework based on a multi-output quantile LSTM architecture for industrial energy management with on-site renewables. It uses conformal prediction to improve prediction interval reliability, generating probabilistic forecasts of PV generation, grid import, and carbon intensity. Evaluated on high-resolution industrial data under winter and summer conditions, it achieves median forecast errors below ~0.7 in winter and ~1.1 in summer, with calibrated prediction intervals. The results demonstrate the suitability of probabilistic deep learning for uncertainty-aware industrial energy planning.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
日本の産業部門では、SSBJ開示やGHG排出削減目標に対応するため、自家発電や再エネ導入が進んでいる。本手法は、太陽光発電や系統電力の不確実性を考慮したエネルギー管理を可能にし、Scope 2排出量の正確な予測と削減計画に貢献する。特に、工場のエネルギー調達最適化やカーボンフットプリント管理において実用的価値が高い。
In the global GX context
Globally, industrial decarbonization requires integrating variable renewables and managing carbon intensity. This probabilistic forecasting framework addresses operational uncertainty in PV generation and grid carbon intensity, supporting real-time energy management and procurement decisions. It aligns with the need for robust forecasting tools under the evolving TCFD/ISSB disclosure landscape, where companies must demonstrate credible transition plans.
👥 読者別の含意
🔬研究者:This paper offers a practical probabilistic deep learning approach for industrial energy forecasting with conformal prediction, relevant to researchers working on uncertainty quantification in low-carbon energy systems.
🏢実務担当者:Corporate sustainability teams can use these methods to improve the reliability of renewable generation forecasts and grid carbon intensity predictions, aiding in Scope 2 reduction strategies and operational planning.
📄 Abstract（原文）
Industrial energy systems integrating on-site renewable generation are subject to increased operational variability and uncertainty, particularly under high renewable penetration and carbon-constrained operation. Conventional deterministic forecasting approaches are insufficient for capturing uncertainty in renewable generation, grid import behaviour, and associated carbon intensity signals. Reliable probabilistic forecasting is therefore required to support next-generation industrial energy management systems. This paper presents a probabilistic multi-horizon forecasting framework based on a multi-output quantile Long Short-Term Memory (LSTM) architecture. The model generates probabilistic forecasts of photovoltaic (PV) generation, grid electricity import, and grid carbon intensity. Prediction interval reliability is improved using conformal prediction, enabling statistically consistent uncertainty calibration without distributional assumptions. The framework is evaluated using high-resolution industrial energy datasets under winter and summer operating conditions. Results demonstrate robust forecasting performance across multiple horizons, achieving median forecast errors below approximately 0.7 in winter and approximately 1.1 in summer. Calibrated prediction intervals achieve empirical coverage levels of approximately 0.80-0.83 in winter and 0.71-0.74 in summer. The results demonstrate the suitability of probabilistic deep learning methods for uncertainty-aware forecasting in industrial energy planning and management.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.1049/icp.2026.2231first seen 2026-06-18 04:56:52
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。