Bayesian network-based resilience assessment of interdependent infrastructure systems under optimal resource allocation strategies

Critical infrastructure systems (CISs) play a key role in the socio-economic activity of a society, but are exposed to an array of disruptive events that can greatly impact their function and performance. Therefore, understanding the underlying behaviors of CISs and their response to perturbations is needed to better prepare for, and mitigate the impact of, future disruptions. Resilience is one characteristic of CISs that influences the extent and severity of the impact induced by extreme events. Resilience is often dissected into four dimensions: robustness, redundancy, resourcefulness, and rapidity, known as the "4Rs". This study proposes a framework to assess the resilience of an infrastructure network in terms of these four dimensions under optimal resource allocation strategies and incorporates interdependencies between different CISs, with resilience considered as a stochastic variable. The proposed framework combines an agent-based infrastructure interdependency model, advanced optimization algorithms, Bayesian network techniques, and Monte Carlo simulation to assess the resilience of an infrastructure network. The applicability and flexibility of the proposed framework is demonstrated with a case study using a network of CISs in Austin, Texas, where the resilience of the network is assessed and a "what-if" analysis is performed.