Typical Scene Acquisition Strategy for VPP Based on Multi-scale Spectral Clustering Algorithm

Abstract

The virtual power plant (VPP) technology can effectively realize the connection of distributed resources and power systems, and is benefit to the safety and economical operation of power systems. This paper considers the value of electric and heat load, fluctuation and changing trend of renewable energy sources such as wind power and photovoltaic. A multi-scale similarity metric method for multi kinds of flexible resources in VPP is proposed, which can be used to measure the similarity of different scenes in a VPP. Based on the proposed multi-scale similarity metric method, the multiscale spectral clustering algorithm is used to obtain typical scenes of VPP. Simulation results show that based on the proposed multi-scale spectral clustering algorithm, considering different kinds of flexible resources, typical scene of VPP can be obtained effectively. And the proposed strategy has satisfying performance and high stability. Also it offers a certain guiding significance for the analysis of VPP flexibility resources, rational allocation of resources and optimization planning.