LAUSR

Abstract Identifying critical post-disaster scenarios in terms of their socioeconomic consequences is of the utmost importance for disaster risk management of infrastructure networks. Recently, multi-objective genetic algorithms have been employed to find such critical post-disaster scenarios for complex systems. However, a large size of a network may hamper multi-objective genetic algorithm from obtaining final solutions that are accurate and robust against variability. To overcome this challenge, a Multi-Group Non-dominated Sorting Genetic algorithm (MG-NSGA) is proposed in this study. It is presented in the paper that MG-NSGA can effectively identify critical post-disaster scenarios by improving the diversity of sample populations. Furthermore, a concept of ‘critical zone’ in the solution space is proposed to determine a group of important post-disaster scenarios identified by MG-NSGA. By large-size real infrastructure network examples of EMA highway network and Jeju transportation network, the proposed MG-NSGA-based identification method is successfully demonstrated.