1. College of Energy and Electrical Engineering, Hohai University, Nanjing 210098, China; 2. Yancheng Power Supply Company of State Grid Jiangsu Electric Power Company, Yancheng 224002, China

Output uncertainties of the wind power, photovoltaic and other renewable energy sources, together with the fluctuation of market price will lead to the risk of the profit of the virtual power plant(VPP). The reasonable allocation of the capacity of wind turbine generators, photovoltaic generation, battery and conventional units can improve the reliability of power supply and maximize the interests of investors. This paper proposes a method of optimizing the capacity of units in VPP considering risk measurement based on the investment portfolio theory that both investment and operation cost are included in the objective. The conditional value-at-risk(CVaR)is set as the risk measurement index, and the impact of risk preference on the multi-energy capacity allocation of VPP is investigated. Historical data of wind, photovoltaic resource and market price in Texas of the United States are employed as the representative scenarios, by using the scenario technologies to simulate uncertainties. The results validate the effectiveness of the proposed model, and provides a quantitative basis for the investors with different risk preferences when planning the multi-energy capacity optimal allocation of VPP problem.