Abstract: Virtual power plant（VPP） can aggregate a large number of distributed energy resources（DER） to participate in the dispatching of power system as a whole. In order to realize efficient control of internal DER in VPP, a virtual power plant cluster autonomous scheduling architecture was proposed. In order to solve the optimization problem of internal DER dynamic tracking scheduling instruction in VPP under ideal communication network, the algorithm of adjusting cost consistency and adjusting time consistency is proposed from the cluster level and DER level respectively. Through the information exchange of local neighboring DER, the parameters of autonomous node are updated, which reduces the communication pressure and improves the solving efficiency. On this basis, by introducing consensus gain function and virtual consensus variable, the transmission delay and noise problems in the real communication network are solved. Simulation results verify the effectiveness and feasibility of the proposed algorithm.