The operation of virtual power plants often faces a large number of uncertain variables, including market prices, new energy power and load. These uncertain variables have caused the virtual power plant to lose load, not allowed to enter the power market or reduce its economic benefits. To avoid these risks, virtual power plants often adopt conservative operational strategies. In order to evaluate the degree of risk brought by the uncertain variables to virtual power plants, this paper first establishes the virtual power plant loss load function, the regulation performance calculation function and the virtual power plant operating profit function. Secondly, based on the Value at Risk method and the three functions mentioned, this paper proposes the conditional value at risk function of the virtual power plant to evaluate the loss of the virtual power plant under extreme risk conditions. Finally, based on the Conditional Value at Risk method, a virtual power plant operation decision model with uncertain risk prevention strategy is proposed. The example shows that the proposed model can effectively reduce the probability of the uncertain risk of the virtual power plant and reduce the loss of the virtual power plant under extreme risk conditions.