Nowadays the energy problems are gradually paid attention, the development of distributed generation(DG) and network communication technology relieves the strain on energy demand, improves the environment problems, and increases the permeability of renewable energy. But up to the present day, the optimization study of the common virtual power plant(VPP) focuses more on the energy scheduling of the internal DGs and the supply of the ancillary services such as market transactions for the external power grid etc. However, the common research fails to consider the load diversity and the electricity flexibility of the user demand, and the initiative and interactivity participating in internal energy scheduling as resources of demand side response. On this basis, the non-cooperative game theory is employed to build the interaction model of the power generation side and the demand side of users in this study. This model provides a theoretical basis to solve problems such as how to improve the interaction flexibility between the power generation side and the demand side of users, considering the satisfaction of users, gaming with their respective maximization of benefit as the goal. The simulation results verify the feasibility of the proposed theory, and compared with the traditional VPP optimization solution the method can realize the economic optimality of both the power generation side and the users' load side considering the users' satisfaction. Moreover, the proposed method improves the utilization rate of renewable energy and realizes the economic optimal operation of the VPP.