The invention discloses a virtual power plant double-layer optimization scheduling method containing an aggregation temperature control load and a new energy. The method includes: (1) a user locally measures and calculates temperature control load power dynamic adjusting external characteristics of each temperature control device in different indoor set temperatures, and obtains aggregation temperature control load power dynamic adjusting external characteristics through aggregation; (2) the aggregation temperature control load power dynamic adjusting external characteristics are put into an upper intra-day rolling optimization scheduling model to update a virtual power plant planned net power curve; (3) the user locally measures and calculates the temperature control load power static adjusting external characteristics, and obtains aggregation temperature control load power static adjusting external characteristics through aggregation; and (4) the aggregation temperature control loadpower static adjusting external characteristics are put into a lower real-time load tracking model for real-time power tracking. According to the method, the maximum unbalanced power of a virtual power plant can be reduced, the influence of the user parameter distribution non-uniformity is avoided, and the method is more applicable to the virtual power plants with various users.