Sequential load restoration with decision-dependent 5G base station backup batteries for resilient distribution systems

-Spare backup batteries of numerous 5G base stations (BSs) can provide considerable flexibility for DS restoration. Meanwhile, their operations are tightly coupled with distribution system (DS) restoration decisions. However, the decision-dependent behaviors of 5G BSs were mostly ignored in previous studies, potentially hindering the DS's secure operation and rapid restoration. To bridge this gap, we have formulated a three-stage model for the operational evolution of 5G BSs. Firstly, backup batteries power BS communication during the outage. In the following two stages, as the DS power is restored, BSs are re-energized, and the backup batteries' spare capacity is leveraged to assist DS load restoration through demand response (DR). Furthermore, a bi-level optimization model is proposed to coordinate sequential load restoration and 5G BS operation, aiming for safer and faster load restoration, with the three evolution stages of 5G BSs seamlessly integrated. This model is solved intractably as nonlinearity terms and non-convex lower-level formulation. Thus, the model linearization is conducted, and a reformulated Column-and-Constraint Generation (R-CCG) method is utilized to handle the fault of Karush-Kuhn-Tucker (KKT) conditions. Numerical results demonstrate that the proposed methodology can accurately capture the decision-dependent characteristic of 5G BS and effectively enhance the DS resilience process.