Deep Projected Gradient Network to Accelerate Low-Carbon Economic Dispatch Considering Energy Storage
エネルギー貯蔵を考慮した低炭素経済運用を加速する深層射影勾配ネットワーク (AI 翻訳)
Qian Ma, Chunxiao Liu, Kui Huang, Qinglin Zou, Zelong Lu, Xianzhuo Liu, Binbin Chen, Jingjing Wang, Zuyi Li
🤖 gxceed AI 要約
日本語
本論文は、カーボン排出取引コスト、風力発電、エネルギー貯蔵を考慮した経済運用問題をリアルタイムで解くため、射影勾配法を深層学習に拡張したD-PGNetを提案する。物理制約を微分可能な射影層として組み込み、最適性ギャップ1.08%未満、カーボン排出偏差0.2%未満、最大64倍以上の高速化を達成した。
English
This paper proposes D-PGNet, a deep projected gradient network that accelerates low-carbon economic dispatch with energy storage and carbon trading costs. It embeds physical constraints as differentiable projection layers, achieving optimality gap <1.08%, carbon emission deviation <0.2%, and up to 64x speedup, enabling real-time decision support.
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
日本の電力システムでは再生可能エネルギー導入拡大に伴い、需給調整やカーボンコストを考慮した経済運用のリアルタイム化が課題。本手法はSSBJ開示やカーボンプライシング政策下での運用最適化に寄与し、電力会社やアグリゲーターの実務に応用可能。
In the global GX context
Globally, the transition to low-carbon power systems requires real-time dispatch solutions that incorporate carbon costs and storage. This work demonstrates a novel AI-based method that could support grid operators and market participants in meeting decarbonization targets while maintaining reliability, aligning with TCFD/ISSB climate risk disclosures.
👥 読者別の含意
🔬研究者:AIと物理制約の融合による最適化手法の新規性と収束性解析に注目。
🏢実務担当者:電力系統運用のリアルタイム最適化とカーボンコスト削減に活用可能。
🏛政策担当者:低炭素電力システムの運用効率向上と排出削減へのAI活用の可能性を示す。
📄 Abstract(原文)
Under the strategic goals of “peak carbon emissions and carbon neutrality”, traditional methods for solving economic dispatch problems involving carbon emission trading costs, wind power, and energy storage devices lack real-time performance and are difficult to support in real-time decision-making. This paper proposes an accelerated solution framework that expands the projection gradient descent method into a Deep Projected Gradient Network (D-PGNet). The network consists of K structured layers, each layer strictly embedding differentiable projection operations corresponding to physical constraints such as power balance, unit ramp-up, and energy storage timing dynamics. This paper systematically derived the projection closed-form solutions of each constraint set to the basic subset, designed an efficient differentiable projection layer based on Dykstra alternating projection, and analyzed the differentiability and convergence properties of the network. Multiple scenario tests have shown that the optimal gap of D-PGNet results is less than 1.08%, the carbon emission deviation is less than 0.2%, the solving speed is improved by more than 64 times at most, and the maximum violation of all physical constraints is below 1.3 × 10−5 p.u.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.3390/pr14152522first seen 2026-08-08 04:57:07
🔔 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。