PVとハイブリッド蓄電システムを備えた同相牽引給電システムの確率制約とCVaRを用いた最適運用
Chance-Constrained and CVaR Optimal Dispatch for Co-Phase Traction Power Supply System with PV and HESS (原題)
Shaofeng Xie, Jingyuan Qi, Hui Wang, Yuqiang Xu, Fan Zhong
🤖 gxceed AI 要約
日本語
本論文は、PVとハイブリッド蓄電システムを統合した同相牽引給電システム（CTPSS）のリスク考慮型最適運用フレームワークを提案する。並列TCN-BiLSTM-Attentionによる確率的PV予測とt-Copulaシナリオ生成を組み合わせ、確率制約計画法とCVaRで経済性と極端リスクを同時に制御する。提案手法は高精度な予測（R2=0.9979）を達成し、許容停電確率0.5%で運用コストを1.36%削減する。
English
This paper proposes a risk-aware optimal dispatch framework for co-phase traction power supply systems (CTPSS) with PV and hybrid energy storage. It integrates probabilistic PV forecasting using parallel TCN-BiLSTM-Attention and t-Copula scenario generation, then applies chance-constrained programming and CVaR to balance economy and extreme risk. The method achieves high prediction accuracy (R2=0.9979) and reduces operating cost by 1.36% at a 0.5% supply loss probability.
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
日本の鉄道セクターでは、カーボンニュートラル目標に向け再生可能エネルギーの活用が進む。本手法は、PV変動下でのリスク管理と経済性の両立を可能にし、鉄道事業者のエネルギーコスト削減と安定供給に寄与する。SSBJ開示や投資家対応の観点からも、再生可能エネルギー統合の実証例として参考になる。
In the global GX context
Globally, this work contributes to the integration of renewable energy in traction power systems, aligning with decarbonization goals in transport. The risk-aware optimization framework offers a methodological reference for managing PV uncertainty in critical infrastructure, relevant to ISSB-aligned climate risk disclosure and transition finance for low-carbon transport.
👥 読者別の含意
🔬研究者:Provides a novel integration of probabilistic forecasting and risk-constrained optimization for renewable-integrated power systems.
🏢実務担当者:Offers a practical scheduling framework for railway operators to reduce costs and enhance renewable utilization.
🏛政策担当者:Demonstrates technical feasibility of low-carbon traction systems, supporting policy for transport decarbonization.
📄 Abstract（原文）
Co-phase traction power supply system (CTPSS) integrated with photovoltaic (PV) and hybrid energy storage system (HESS) can enhance power supply capacity while facilitating the on-site consumption of renewable energy. This integration provides a pathway toward green development and the achievement of the “carbon peaking and carbon neutrality” goal. However, the stochastic nature of PV poses challenges to safe and economical operation of the system. Existing energy management methods remain limited in simultaneously balancing operational economy and extreme risks caused by PV uncertainty, making it difficult to achieve an effective trade-off between operating cost and constraint violation risk. To address this issue, a risk-sensitive optimal scheduling framework integrating probabilistic PV forecasting, correlated scenario generation, and risk-aware optimization is developed. First, a PV probabilistic prediction model based on parallel TCN-BiLSTM-Attention is proposed, which extracts multiscale local features and long-range temporal features from PV for accurate uncertainty quantification. Second, a t-Copula PV scenario generation method driven by weather classification and temporal correlation is proposed. On this basis, a day-ahead optimal scheduling strategy integrating chance constraints programming (CCP) and conditional value at risk (CVaR) is established to simultaneously control constraint violation risk and extreme economic risk. The proposed prediction model can accurately quantify uncertainty, achieving an R2 value of 0.9979. When the allowable power supply loss probability is 0.5%, the proposed scheme’s operating cost is reduced by 1.36% compared with the baseline scheme. The results demonstrate that the proposed framework can effectively coordinate operating economy and extreme-risk control, thereby improving the risk-sensitive operational performance of CTPSS with PV and HESS. The proposed strategy balances economy and extreme risk, providing a reference for safe, low-carbon and economical operation of CTPSS with PV and HESS.
🔗 Provenance — このレコードを発見したソース
- primary_source https://doi.org/10.3390/wevj17090470first seen 2026-09-07 00:12:29
🔔 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。