Two-stage optimization scheduling strategy for virtual power plants considering the correlated uncertainty of distributed energy storage in extreme scenarios

Under extreme scenarios, virtual power plants (VPPs) are subjected to multiple impacts such as abrupt changes in source-load power and mismatched coordinated operation of distributed energy storage (DES) clusters, which significantly increases the difficulty of balancing scheduling security and economic efficiency. Considering the correlated uncertainty of DES, this article proposes a two-stage optimization scheduling strategy for VPPs to address the issue of enhancing system resilience in extreme disaster events. Firstly, K-means clustering is performed on the wind and load data, and based on typical scenario screening, an extreme scenario optimization selection mechanism is proposed with the principle of maximizing scenario dissimilarity. Secondly, based on cross correlation analysis, the correlated uncertainty of DES power changes in extreme scenarios is characterized, and a correlated matrix is constructed. Then, the two-stage optimization scheduling strategy is proposed. In the first stage, the total cost of purchasing electricity and operation and maintenance expenses is minimized. In the second stage, based on the output of each DES obtained in the first stage, the correlated uncertainty between DES and their own response deviation are reduced. Finally, the effectiveness of the strategy is verified through numerical analysis. The results show that compared to existing methods, the proposed strategy can reduce 30.91% of the power gap and 68.77% of the total cost in extreme scenarios of the distribution network. Therefore, the proposed method effectively balances the risk resistance ability and scheduling economy in extreme scenarios, providing a new technical path for the safe and efficient operation of VPP under extreme working conditions.