An Approach to Virtual Power Plants Multi-Level Aggregation Considering Distributed Resource Characteristics

As a crucial mechanism for coordinating and optimizing distributed energy resources (DERs), the aggregation capability of virtual power plants (VPPs) significantly influences both market participation effectiveness and power system operational efficiency. This paper proposes a multi-level aggregation methodology for VPPs that explicitly accounts for the inherent characteristics of diverse DERs. Initially, we establish comprehensive mathematical models for heterogeneous DERs based on their distinctive operational characteristics. Subsequently, we construct a two-level aggregation framework leveraging Zonotope theory: at the DER's level, the framework employs generator matrices to capture the dynamic characteristics of homogeneous resources, enabling characteristic-preserving aggregation within each resource category; at the VPP level, the overall optimization of heterogeneous resources is aggregated through the minimum volume deviation to ensure the accuracy and operability of the aggregation results, and further based on the vertex search method to solve the impact of network constraints. Case analysis shows that compared with the traditional methods, the proposed method effectively reduces the model complexity while maintaining the aggregation accuracy, and realizes the accurate and fast portrayal of the feasible domain of VPP aggregation.