A novel approach for resilience optimization and post-decision assessment considering limited recovery resource

For Critical Infrastructures (CIs), resilience has emerged as a critical issue in assessing their ability to withstand and recover from disruptive events. However, most of the existing work overlooks the integration of optimization policies and post-decision assessment, leaving a gap in understanding the long-term effectiveness of resilience-enhancing decisions. In this paper, we present a comprehensive framework for optimizing resilience and a methodology for post-decision evaluation. For a deteriorating system with multiple subsystems, we explicitly model the relationships between system state and performance. The maximization of the expected discounted long-run performance under a limited intervention resource is taken as the optimization objective. The problem is formulated within a constrained Markov decision process (CMDP) framework. The system resilience is fully assessed afterwards: both traditional metrics and novel metrics are explored. Through numerical studies and a case study, the effectiveness of this work is presented. The results demonstrate that the recovery budget has a significant impact on system performance and other resilience assessment measures. It offers valuable managerial insights into the allocation of the recovery budget and the evaluation of post-decision outcomes.