Extreme natural disasters pose a serious challenge to power system reliability, often causing widespread outages and service disruptions. Networked microgrids—interconnected clusters of distributed energy resources and loads—require sophisticated restoration strategies that coordinate multiple objectives: identifying and repairing faults quickly, restoring customer loads in priority order, reconfiguring networks to isolate damage, and scheduling limited repair crews efficiently.
Traditional post-disaster restoration approaches typically handle fault repair and load restoration sequentially, which delays recovery. Researchers have now proposed a co-optimization framework that treats these processes as interdependent rather than separate problems.
The method begins by modeling the fault repair challenge as a multi-objective problem balancing repair time, economic loss, labor costs, and crew efficiency. A key innovation is the adaptive quantum particle swarm optimization algorithm, which employs nonlinear contraction and expansion coefficients to navigate the solution space more effectively than conventional algorithms. This approach improves convergence speed and solution quality—critical factors when minutes of outage translate to significant costs and public safety risks.
The proposed two-layer co-optimization model then integrates these repair decisions with load restoration and network reconfiguration strategies. As repairs progress, the algorithm dynamically updates the network topology, ensuring that restored areas are connected safely and that loads are restored in a priority sequence that maximizes grid stability and customer service.
Validation used modified IEEE 33-bus and PG&E 69-bus test networks, representing realistic medium-sized distribution systems. Results demonstrated 15.82% improvement in restoration speed and 6.58% better resilience in load retention capacity compared to existing methods—meaningful gains for utilities managing disaster response.
As climate change increases extreme weather frequency and severity, rapid restoration capabilities become essential for grid resilience. This research provides utility operators with a practical computational tool to minimize outage impacts and accelerate recovery in networked microgrid environments.
