Scholarlink

<p>Vehicle-to-Grid (V2G) services are a promising source of flexibility in decarbonized power systems, requiring the coordination of numerous geographically dispersed electric vehicles (EVs) to balance grid operator and user charging demands. This poses complex decision-making challenges for the Electric Vehicle Aggregator (EVA) due to significant and varied uncertainties. These challenges result in large-scale, high-cardinality problems that make direct, high-granularity resource allocation computationally infeasible and difficult to analyze. To address this, the paper proposes a reduced-order, multi-stage framework to support the EVA in providing V2G services in the day-ahead market and scheduling individual EVs throughout the day. The framework introduces several innovations. It integrates a deterministic EV scheduling model into a scenario generation tool to estimate V2G capacity across different scenarios over time. A reduced-order stochastic optimization approach is then used to tackle the day-ahead market bidding problem, significantly lowering computational complexity. Additionally, during the intra-day phase, the EVA employs a novel strategy to manage imbalances between contracted and actual charging/discharging profiles, leveraging these imbalances to boost profitability while considering traditional penalty schemes. Numerical simulations using real market data validate the framework’s effectiveness by analyzing intra-day scenarios across various timeframes.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multi-Stage Decision Framework for EV Aggregators in Day-Ahead Markets using Low-Granularity Power Profiles Scenarios

  • 期刊论文

  • 发表日期：

    2026年2月26日

查看全文
  • Fabrizio De Caro,
  • Giuseppe Graber,
  • Vito Calderaro,
  • Alfredo Vaccaro,
  • Vincenzo Galdi

摘要

Vehicle-to-Grid (V2G) services are a promising source of flexibility in decarbonized power systems, requiring the coordination of numerous geographically dispersed electric vehicles (EVs) to balance grid operator and user charging demands. This poses complex decision-making challenges for the Electric Vehicle Aggregator (EVA) due to significant and varied uncertainties. These challenges result in large-scale, high-cardinality problems that make direct, high-granularity resource allocation computationally infeasible and difficult to analyze. To address this, the paper proposes a reduced-order, multi-stage framework to support the EVA in providing V2G services in the day-ahead market and scheduling individual EVs throughout the day. The framework introduces several innovations. It integrates a deterministic EV scheduling model into a scenario generation tool to estimate V2G capacity across different scenarios over time. A reduced-order stochastic optimization approach is then used to tackle the day-ahead market bidding problem, significantly lowering computational complexity. Additionally, during the intra-day phase, the EVA employs a novel strategy to manage imbalances between contracted and actual charging/discharging profiles, leveraging these imbalances to boost profitability while considering traditional penalty schemes. Numerical simulations using real market data validate the framework’s effectiveness by analyzing intra-day scenarios across various timeframes.