A resilience-oriented optimization framework for smart grid operation and recovery before, during, and after natural disasters

The increasing frequency of natural disasters driven by climate change poses significant challenges to the resilience of power distribution networks. Enhancing grid resilience requires proactive measures before disruptions, efficient reconfiguration during events, and strategic recovery planning post-event. This paper presents a novel three-stage optimization framework aimed at fortifying distribution networks against natural disasters. The proposed model integrates proactive asset positioning, Dynamic Feeder Reconfiguration (DFR), Mobile Emergency Generator (MEG) deployment, and Direct Load Control (DLC) to minimize Expected Energy Not Served (EENS) and forced load shedding. In the first stage, Automatic Switches (ASs) are optimally placed to enhance system flexibility. The second stage focuses on the pre-positioning of MEGs and Crew Teams (CTs) for rapid emergency response. The final stage addresses generation scheduling and repair strategies to accelerate post-disaster recovery. The optimization is formulated as a Mixed-Integer Linear Programming (MILP) model and solved using CPLEX in GAMS. Case studies on a modified IEEE 69-bus distribution network demonstrate that the proposed strategy reduces EENS by 43.89 % and forced load shedding by 64.05 %, ensuring greater system reliability and faster recovery. These results underscore the effectiveness of integrated resilience strategies for smart grid operations under extreme events.