Distributionally robust optimal scheduling of flexible distribution networks considering dynamic spatio-temporal correlation of renewable energy

The integration of renewable energy sources (RESs) at the end of the grid has substantially increased system uncertainties, thereby complicating scheduling processes and highlighting the urgent need for the establishment of efficient flexible distribution networks (FDNs). Existing methods often treat spatio-temporal correlations among RESs in a static manner, neglecting their dynamic variations and resulting in suboptimal dispatch strategies. To address these challenges, this study proposes a distributionally robust optimal scheduling strategy that incorporates dynamic spatio-temporal correlations of RESs based on Copula functions and Markov theory. The proposed approach establishes an optimal power flow model for FDNs, considering the synergistic effects of various flexible regulators resources such as Flexible Multi-State Switches with Hydrogen Energy Storage (H-FMSS), Static Var Generator (SVG), Capacitor Bank (CB), and Flexible Load (FL). By utilizing the 1-norm and ∞-norm to characterize the uncertainty sets between sources and loads, this strategy effectively minimizes operational costs while ensuring robust performance. The proposed model is linearized and relaxed through second-order cone programming and solved using the column and constraint generation (C&CG) algorithm. Case studies utilizing the improved IEEE 33-bus and PG&E 69-bus system demonstrate that the proposed strategy offers a more accurate and resilient scheduling method for FDNs.