レジーム依存不確実性下における検索拡張型ノンパラメトリック確率的风力発電予測
Retrieval-Augmented Non-Parametric Probabilistic Wind Power Forecasting under Regime-Dependent Uncertainty (原題)
Sun¹ H, Guo¹ S, Pan² X, Qu¹ Z, Ma¹ J, Xu¹ Y
🤖 gxceed AI 要約
日本語
本論文は、物理情報を組み込んだBiLSTM–Attentionと検索拡張型ノンパラメトリック推定を組み合わせ、風力発電の超短期確率予測手法を提案する。深層特徴・風速近接性・時間連続性を統合した近傍重み付けと拡散平滑化カーネル密度推定により、点予測と予測区間を同時に生成する。浙江と内蒙の2風力発電所での実験で、Transformer比でMAE・RMSEを約2割改善し、予測の追跡可能性も示した。
English
This paper proposes a retrieval-augmented non-parametric framework for ultra-short-term probabilistic wind power forecasting. A physics-informed BiLSTM-Attention network extracts regime features, and retrieved historical analogs are weighted by deep-feature similarity, wind-speed proximity, and temporal continuity before a diffusion-smoothed kernel density estimator produces point forecasts and prediction intervals. On two Chinese wind farms, it cuts MAE/RMSE by ~21-23% versus a Transformer baseline and improves the 80% Winkler score by 24.4% versus CKDE, while making forecasts traceable to historical samples.
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
再エネ大量導入に伴う系統安定化と出力予測高度化は日本のGX政策(再エネ主力電源化・需給調整市場)と直結する。予測区間の精度向上は発電事業者の需給計画やインバランス回避に寄与しうる。
In the global GX context
Accurate probabilistic renewable forecasting underpins grid integration and is increasingly relevant to transition finance and climate-risk modeling as variable renewables scale. The retrieval-based traceability also aligns with growing demand for explainable AI in energy and disclosure contexts.
👥 読者別の含意
🔬研究者:検索拡張と拡散平滑化KDEを組み合わせた確率予測の設計は、再エネ予測研究の新たなベンチマークになりうる。
🏢実務担当者:風力発電事業者は予測区間の精度向上により需給計画やインバランスコスト管理に活用できる。
🏛政策担当者:再エネ統合政策や系統運用ルール設計において、高精度な確率予測の活用可能性を示す。
📄 Abstract(原文)
<title>Abstract</title> <p>Reliable ultra-short-term probabilistic wind-power forecasting is important for renewable-energy integration, especially during ramps and extreme fluctuations. Existing deep probabilistic models typically encode historical operating regimes only implicitly, while conventional non-parametric estimators rely on low-dimensional similarity measures and coarse sample weighting. This paper proposes a retrieval-augmented non-parametric framework that reconstructs the conditional distribution of wind power from physically relevant historical analogs. A physically informed BiLSTM–Attention network extracts regime features from meteorological and historical power inputs; candidate historical samples are then retrieved and weighted by combining deep-feature similarity, wind-speed proximity, and temporal continuity, and the weighted neighbors are passed to a diffusion-smoothed kernel density estimator that produces both point forecasts and prediction intervals. Experiments on two heterogeneous wind farms in Zhejiang and Inner Mongolia show that, on Dataset A at the 60-min horizon, the method lowers MAE and RMSE by 21.2% and 22.7% relative to a Transformer baseline and reduces the 80% Winkler score by 24.4% relative to CKDE. A retrieved-neighbor analysis further makes each forecast traceable to physically meaningful historical samples.</p>
🔗 Provenance — このレコードを発見したソース
- Research Square https://doi.org/10.21203/rs.3.rs-11387551/v1first seen 2026-11-13 04:30:40
🔔 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。