Distributionally Robust Optimal Dispatch of a Rural Integrated Energy System Under Extreme Cold-Wave Weather
極寒波気象下における農村統合エネルギーシステムの分布ロバスト最適配分 (AI 翻訳)
Liu L, Gong G, Li Z, Yang J, Yang C
🤖 gxceed AI 要約
日本語
極寒波下の農村統合エネルギーシステム（RIES）に対し、条件付き変数時間生成的敵対ネットワーク（CVT-GAN）を用いた分布ロバスト最適化手法を提案。太陽光出力や熱負荷のシナリオ生成とWasserstein ambiguity setにより、最悪ケースでの多エネルギー協調配分を実現。電気・熱回復率が向上し、コストも低減。
English
This paper proposes a distributionally robust optimization method for rural integrated energy systems under cold-wave weather, using a conditional GAN (CVT-GAN) to generate scenarios of PV output, heat load, and temperature. A Wasserstein ambiguity set is built around these scenarios, and the multi-energy dispatch is optimized for worst-case distribution. Case studies show improved recovery rates and cost reduction.
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
日本の寒冷地域（北海道・東北）における農村エネルギーシステムのレジリエンス向上に示唆を与える。再生可能エネルギー変動と厳冬期の熱需要逼迫への対応策として、SSBJやTCFDが求める気候リスク管理にも資する。
In the global GX context
This study offers a method for enhancing energy resilience under extreme cold, relevant to ISSB's climate resilience disclosures and global adaptation strategies. The use of AI-generated scenarios for worst-case dispatch aligns with climate risk quantification demands.
👥 読者別の含意
🔬研究者:Provides a novel integration of CVT-GAN with distributionally robust optimization for energy dispatch under extreme weather.
🏢実務担当者:Offers a practical dispatch strategy for integrated energy systems to maintain critical loads during cold waves.
🏛政策担当者:Highlights the importance of robust energy infrastructure planning for climate adaptation.
📄 Abstract（原文）
Cold-wave events increase heat-supply pressure, renewable energy fluctuations, and greenhouse environment maintenance difficulty in a rural integrated energy system (RIES), threatening rural livelihoods and agricultural production. To enhance optimal dispatch and critical load support under cold-wave conditions, this paper proposes a Wasserstein distributionally robust optimisation (WDRO) dispatch method that uses a conditional variable temporal generative adversarial network (CVT-GAN) to generate scenarios. CVT-GAN first generates multivariate cold-wave scenarios involving photovoltaic (PV) output, residential heat load, and outdoor temperature. A Wasserstein ambiguity set is then constructed around the empirical distribution of these scenarios, and the multi-energy coordinated dispatch strategy of the RIES is optimised under the worst-case distribution. The original min-sup model is reformulated as a single-level mixed-integer linear programming (MILP) model. Case studies show that, under a staggered joint electricity–gas interruption scenario, the electric and heat recovery rates reach 0.8608 and 0.9604, respectively. Compared with the average results of deterministic optimisation (DO) and stochastic programming (SP), the recovery capabilities increase by 0.29% and 2.42%; compared with robust optimisation (RO), the expected cost decreases by 28.42%, and the recovery rates increase by 5.35% and 5.19%.
🔗 Provenance — このレコードを発見したソース
- Research Square https://doi.org/10.22541/authorea.15004397/v1first seen 2026-06-07 04:22:59 · last seen 2026-06-16 04:30:24
🔔 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。