基于矩不确定性的虚拟发电厂分布鲁棒机组组合OACSTPCD

Due to the randomness of output from distributed energy resources,it is difficult for a virtual power plant(VPP)to provide specific physical parameters to independent system operators(ISOs).Therefore,a distributionally ro-bust optimization(DRO)model based on higher-order moment ambiguity sets is established in this paper to evaluate the physical characteristics of the VPP.First,the moment information(e.g.,mean and covariance)of virtual net load is used to describe uncertainties in wind power output and load demand.Then,the unit commitment of the VPP is opti-mized at the first stage,the unit output and energy storage charge and discharge are optimized at the second stage,and a two-stage DRO model of the VPP is established.Finally,combined with dual transformation,the distributionally ro-bust unit commitment model of the VPP is transformed into a quadratic second-order cone programming model,which is convenient to be solved through two-stage iterations.Through the analysis of examples,the validity of the distribution-ally robust unit commitment model for VPP established in this paper is verified.