Abstract: The construction of a new power system with new energy as the main body needs to deeply explore the scheduling potential of flexible resources on the demand side, so that they can participate in grid scheduling and adjustment to improve grid balance and adjustment capabilities. The characteristics of massive and heterogeneous flexible resources increase the complexity of resource aggregation and solving feasible region. In this regard, based on the vertex search method, this paper establishes a virtual power plant dynamic feasible region solution model that takes into account the uncertainty of distributed power output, and then decouples the internal coupling constraint relaxation of the solution model into a virtual queue stability problem through Lyapunov optimization. A mathematical model aiming at vertex search and queue stability is proposed, and finally a solution algorithm is formed by establishing a virtual time queue. The proposed algorithm improves the solution efficiency by decoupling the coupling constraints, and improves the resource scheduling capability by satisfying the scheduling capability constraints. The example results show that the proposed algorithm has the advantages in improving computational efficiency and constraint satisfaction ability in day-ahead and real-time scenarios.