Distributionally robust resilience optimization of post-disaster power system uncertainties

In an era where extreme weather events are becoming more frequent and severe, the resilience of power systems against such disruptions is vital for societal stability. This study introduces a comprehensive framework for reducing the resilience loss of power systems after such disruptive events, incorporating a detailed analysis of the inherent uncertainties that challenge post-disaster restoration efforts. We categorize these uncertainties into time and demand-related factors and establish a tailored resilience measure to evaluate the efficacy of power system restoration schedules. We develop a two-stage stochastic programming model that minimizes expected resilience loss, integrating the routing of restoration crews-a crucial aspect that directly influences restoration timeliness and efficiency. Furthermore, we pioneer a distributionally robust optimization model utilizing an ambiguity set based on Wasserstein distance to navigate demand uncertainties. The applicability and effectiveness of the proposed models are demonstrated through a case study of Guangxi Province's power grid, illustrating their potential to improve post-disaster recovery strategies.