Virtual power plant （VPP） aggregates distributed power and load-side adjustable resources participating in energy market trading, which has become an effective way to absorb new energy and tap the demand response potential. Under this background, a VPP dispatching model based on multi-objective and two-stage programming is constructed. In the day-ahead stage, the maximization of market operation income and user energy absorption satisfaction has been taken as the goal to construct the multi-objective optimization model of VPP. The Pareto optimal solution of the multi-objective problem is solved by NSGA-‍Ⅱ algorithm, and the comprehensive benefit index is proposed to select the best solution. In the real-time stage, the minimization of deviation penalty cost and wind abandonment cost has been taken as the goal to construct the robust programming model of VPP, considering the influence of the flexible thermoelectric ratio of combined heat and power （CHP） plant on the wind power absorption. Finally, an actual VPP is analyzed to verified the multi-objective two-stage programming model. The results show that the proposed model can improve its energy utilization efficiency on the premise of ensuring the comprehensive benefits of VPP.