Intelligent Rolling Optimization Algorithm for Day-Ahead, Intra-Day, and Real-Time Scheduling of Virtual Power Plants

The integration of Virtual Power Plants (VPPs) into modern power systems requires advanced algorithms for efficient scheduling across multiple time horizons, including day-ahead, intra-day, and real-time markets. This paper proposes an intelligent rolling optimization algorithm designed to enhance the operational efficiency and economic performance of VPPs. The algorithm combines predictive modeling, rolling optimization, and adaptive adjustments to dynamically optimize scheduling decisions while accounting for uncertainties in renewable energy generation, market prices, and demand fluctuations. By leveraging machine learning-based forecasting models and robust optimization techniques, the proposed method ensures the seamless integration of day-ahead plans with intra-day and real-time operations. Experimental results demonstrate the algorithm's effectiveness in improving scheduling accuracy, reducing operational costs, and enhancing system reliability under dynamic market conditions. This work provides a scalable and practical solution for intelligent VPP control in evolving electricity markets.