Robustness-based optimization model for post-flood road network recovery

Abstract

Road networks are often exposed to flood events, causing significant infrastructure damage, economic losses, and mobility disruptions. For an effective post-flood recovery, optimal decision-making is required to repair damaged roads under a minimized budget while maintaining a robust road network. This study develops a network robustness-oriented optimization model for post-flood recovery of road infrastructure. The objective is to identify the minimum-cost subset of road links while maintaining the post-repair network robustness at a given threshold, using the network robustness as the measure of accessibility. The study presents the problem as a binary combinatorial optimization and solves it using a genetic algorithm (GA), which enables operating complex, non-linear constraints subjected to network connectivity. A set of synthetic road networks with rational cost parameters was tested, varying the damage levels and target robustness values. The results show that the model can successfully identify the cost-effective repair plans while restoring network robustness. This framework can provide insight for transportation agencies in the planning of resilient and cost-effective road recovery strategies after disaster events.