A Multi-Time-Scale AGC Frequency Regulation Optimal Scheduling Method for Virtual Power Plants Based on PPO Algorithm

With the rapid development of distributed energy resources and electric vehicles, the virtual power plant (VPP) has attracted significant attention as an effective means to aggregate various distributed resources for grid scheduling. This paper focuses on a VPP consisting of multiple types of resources, including photovoltaic (PV), wind power, gas turbines, energy storage systems (ESS), and electric vehicles (EVs), and establishes a multi-time-scale AGC frequency regulation optimal scheduling model. The model fully considers the dynamic characteristics and complementarity of different resources, and achieves day-ahead economic optimization and intra-day frequency regulation performance enhancement through hierarchical rolling scheduling. To address the problems of traditional scheduling algorithms, such as being prone to local optima and difficulty in balancing economic efficiency and flexibility, this paper proposes an optimization approach based on the Proximal Policy Optimization (PPO) algorithm. Simulation results demonstrate that the proposed method effectively coordinates the outputs of all dispatchable resources, improves the comprehensive frequency regulation performance and operational benefits of the VPP, and provides strong support for building a secure and efficient power system with a high penetration of renewable energy.