Research on the Optimization Scheduling Strategy of Virtual Power Plants Based on Conditional Value at Risk and Random Scenario Screening

Virtual Power Plant (VPP) has become an effective way to improve the stability of power grid and the utilization rate of renewable energy because of its ability to aggregate various distributed energy sources into a whole. However, significant risk losses are faced by the VPP as it seeks to maximize income, owing to the fluctuations in wind power generation and photovoltaic output. A multi-source joint optimal scheduling model of virtual power plant based on Conditional Value at Risk (CVaR) and stochastic optimization is proposed. CVaR risk control is used to reduce the risk caused by fluctuation, and Latin hypercube sampling (LHS) and K-medoids clustering algorithm are combined to optimize the scene with uncertain wind and scenery, so as to improve the scheduling accuracy. The outcomes of the simulation indicate that an effective balance between benefits and risks can be achieved by the model under varying risk preferences, and improve the economy of VPP and the wind power consumption rate.