Importance estimation considering interdependencies and correlations among multiple types of infrastructure

The resilience and functionality of critical infrastructure systems are interconnected and can be affected by both similar and diverse types of infrastructure systems. This paper introduces a novel approach for estimating the importance of individual infrastructure components by considering their interdependencies with both single- and multi-type of infrastructure networks and associated correlations. This approach enables more rational management of multi-type of critical infrastructure networks under uncertainty. The interdependencies between infrastructure components and networks are modeled using a Bayesian network. Subject to normal operational and seismic conditions, the probabilities of failure of the individual infrastructure components are treated as random variables, and the correlation coefficient between the structural performances of components is computed. Given that components with higher correlation coefficients are considered more critical, the importance factors for all infrastructure components are then estimated. The comparison between the importance factors of infrastructure components with and without considering the interdependency with other infrastructure networks is investigated. The integration of these importance factors is presented. Additionally, the importance factors under normal operational and seismic conditions are compared. The proposed approach is illustrated using existing critical infrastructure networks including electric power supply, water supply, telecommunication, and bridge networks.