Post-Blackout Restoration in Distribution Networks via Dynamic Microgrid and Non-Black-Start Units

With the increasing deployment of distributed generators (DGs) in active distribution networks, self-restoration after blackouts has become technically feasible. However, in large-scale blackout scenarios, limited black-start capability and fixed line repair sequences pose significant challenges to efficient and rapid system restoration. To address these challenges, this paper proposes a coordinated restoration strategy that leverages topological flexibility and prioritizes the activation of high-capacity non-black-start units. A mixed-integer linear programming model is formulated to jointly optimize microgrid formation and load restoration across multiple repair stages. The model incorporates fictitious network flows and spanning forest constraints to guide DG activation and system reconfiguration. The proposed method is validated on a modified IEEE 123-node test system. Comparative analysis and Monte Carlo simulations demonstrate that the proposed strategy restores more load in early stages, activates more DGs, and achieves higher DG capacity utilization under diverse outage scenarios.