In order to solve the problem of internal power uncertainty in virtual power plant (VPP), promote the consumption of new energy power and increase users' enthusiasm for participating in system dispatch, firstly, through the establishment of a model, the power uncertainty of each unit in VPP was analyzed. Then, a risk penalty mechanism for wind and solar abandonment and a demand response risk penalty mechanism were established to solve the problems of the power uncertainty of wind and solar generators and demand response. Based on the risk penalty mechanism and combined with the output characteristics of each unit, a VPP coordinated dispatch strategy was formulated. Considering the risk penalty cost and taking into account the system operation constraints, with the maximum net profit as the objective function, a VPP optimal scheduling model was established. Finally, the IEEE 30 node system was selected for simulation and analysis. The scenario analysis method was used to study the impact of the risk penalty mechanism on the optimal scheduling of VPP. The results show that the risk penalty mechanism for abandoning wind and solar energy can promote the consumption of wind and solar energy, and the demand response risk penalty mechanism can increase the enthusiasm of users to participate in scheduling. It can be seen that the establishment of the risk penalty mechanism is effective.