The technology of virtual power plant (VPP) provides an effective approach of aggregating distributed energy and user-side resources to participate in energy trading in energy market. Based on the background of VPP’s multi-energy aggregators participating in demand response programs. A two-stage optimal operation model of VPP is proposed, which integrates energy storage units, energy conversion equipment (CHP, electric boiler, gas turbine, etc.), distributed renewable energy generation, multi-energy trading and incentive-based integrated demand response (IDR). In the first stage, a optimal economic dispatch model of VPP is established. In the second stage, a bi-level IDR model which contains consumer model and aggregator model is formulated, coupled with solution methodology given. Simulation results show merits of the proposed algorithm in reducing the scheduling cost of VPP, the compensation cost of aggregators and the loss of consumer’s comfort.