gxceed
← 論文一覧に戻る

Federated Safe Proximal Policy Optimization for Robust Low-Carbon Dispatch of Heterogeneous Multi-Park Electricity–Heat–Hydrogen Integrated Energy Systems

フェデレーテッドセーフ近位ポリシー最適化による不均一マルチパーク電気・熱・水素統合エネルギーシステムの頑健な低炭素運用 (AI 翻訳)

Zijie Peng, Xiaohui Yang, Qianhua Xiao

Energies📚 査読済 / ジャーナル2026-05-15#エネルギー転換Origin: CN
DOI: 10.3390/en19102382
原典: https://doi.org/10.3390/en19102382

🤖 gxceed AI 要約

日本語

本論文は、不均一なマルチパーク電気・熱・水素統合エネルギーシステム（EHHS）の低炭素運用を目的とし、フェデレーテッドセーフ強化学習に基づく枠組みを提案する。段階的炭素取引とピアツーピアエネルギー取引を考慮し、プライバシー保護と安全性制約を満たすために、制約付きマルコフ決定過程として定式化し、FedSafePPOアルゴリズムを開発した。産業、商業、住宅の3つの異種パークでのケーススタディにより、総運用コストと炭素排出量の削減、学習安定性の向上を実証した。

English

This paper proposes a low-carbon dispatch framework for heterogeneous multi-park electricity-heat-hydrogen integrated energy systems (EHHS) using federated safe reinforcement learning. It incorporates stepped carbon trading and peer-to-peer energy trading, formalized as a constrained Markov decision process. The developed FedSafePPO algorithm integrates PPO, Lagrangian safety handling, and federated aggregation, enabling privacy-preserving and safe dispatch. Case studies with industrial, commercial, and residential parks show reduced costs and emissions with improved training stability.

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

中国の研究だが、段階的炭素取引と水素統合エネルギーシステムの運用最適化手法は、日本のGX政策やSSBJの情報開示枠組みにおけるエネルギー転換の実践に示唆を与える。特に、プライバシー保護を考慮した分散型運用は、日本の地域エネルギー管理に応用可能。

In the global GX context

This study advances global GX research by presenting a novel federated safe RL method for low-carbon dispatch with carbon trading and hydrogen integration. It addresses critical challenges of privacy and safety in multi-agent energy systems, relevant for TCFD/ISSB-aligned disclosure on transition risk and operational decarbonization.

👥 読者別の含意

🔬研究者:A novel algorithmic contribution combining federated learning, safe RL, and carbon trading for energy dispatch, offering a foundation for further research in distributed low-carbon operations.

🏢実務担当者:Energy companies and system operators can explore this method for cost-effective and low-carbon dispatch while preserving operational data privacy.

📄 Abstract（原文）

To achieve low-carbon and cost-effective operation of multi-park electricity–heat–hydrogen integrated energy systems (EHHSs), this paper proposes a low-carbon dispatch framework based on federated safe reinforcement learning. First, a multi-park EHHS dispatch model is established by considering heterogeneous park characteristics, electricity–heat–hydrogen coupling, stepped carbon trading, and peer-to-peer (P2P) energy trading. Then, to address the coupled challenges of privacy preservation, operational coupling, and safety constraints, the dispatch problem is formulated as a constrained Markov decision process (CMDP). On this basis, a federated safe proximal policy optimization algorithm (FedSafePPO) is developed by integrating PPO, Lagrangian-based safety constraint handling, and federated parameter aggregation. The proposed method enables each park to learn a local dispatch policy from private data while sharing global knowledge without exchanging raw operational data. In addition, an actor–dual-critic architecture is adopted to jointly evaluate economic returns and constraint costs, thereby improving convergence stability and dispatch feasibility. Case studies involving three heterogeneous parks—industrial, commercial, and residential—demonstrate that the proposed method effectively reduces total operating costs and carbon emissions while satisfying system constraints. Compared with PPO, FedPPO, and SafePPO, the proposed FedSafePPO achieves superior low-carbon economic performance, greater training stability, and better adaptability to heterogeneous operating conditions. The results verify the effectiveness and engineering applicability of the proposed method for the low-carbon dispatch of multi-park EHHSs.

🔗 Provenance — このレコードを発見したソース

🔔 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。