A Two-Stage Robust Optimization Framework for Community Resilience Planning
Abstract
Community resilience planning is challenging not only because the community is composed of many interdependent infrastructures, a diverse population, and numerous stakeholders but also because the hazards the community is vulnerable to are uncertain, and the availability of resources for mitigation actions is limited. Moreover, communities need to plan for hazards that are very rare but highly consequential, and the level of conservativeness in decision-making significantly impacts the resulting resilience metrics. This work presents a two-stage robust optimization framework that can assist communities in developing mitigation strategies for the community’s built environment and investigate the effect of varying the level of conservativeness in decision-making in terms of several resilience metrics. The first stage model of the proposed framework identifies pre-hazard mitigation actions for the community’s built environment to maximize the resilience of the least favorable hazard scenario among the scenarios under consideration. On the other hand, the second stage model generates the worst hazard scenario in terms of the resilience metric for the current set of actions and the uncertainty sets representing the stochastic nature of the hazard. A case study is presented to showcase the framework's applicability in developing alternative mitigation strategies under varied levels of conservativeness.