This paper focuses on aggregating decentralized demand-side resources through virtual power plants(VPPs) to enhance the peak load management capability in the construction of new electric power systems. The research centers on the core issue of “load-based” virtual power plants participating in power source planning. It investigates the modeling methods for the adjustable potential of virtual power plants

including electric vehicles

air conditioning loads

and industrial loads. Moreover

the paper delineates the economic and operational technical constraints of virtual power plants in power source planning and constructs a power source planning model that accounts for VPPs. Considering the prediction errors of renewable energy and the characterization errors of the VPPs' response potential

the paper utilizes interval optimization theory to reformulate the model into an interval planning model and completes the deterministic transformation of the model through interval order relations and interval possibility. The rationality of the proposed power source planning method is verified through a case study.The results show that the proposed method can effectively aggregate VPP resources

significantly enhancing the system's peak load management capability. The constructed interval planning model can effectively handle forecast errors

ensuring the reliability of planning. This method provides an economically feasible solution for the transformation of the source-load structure in new electric power systems. It is evident that the aggregation and optimized dispatch of VPPs can significantly enhance the flexibility and reliability of the power system.