Resilience optimization strategy for urban transportation networks under higher-order interactions

Abstract Urban metro-bus networks are highly vulnerable to cascading failures under extreme disruptions, yet most recovery models inadequately capture their complex interdependencies. Moving beyond conventional single-mode or static approaches, this paper introduces a higher-order interaction perspective to model the non-linear, synergistic effects in composite transportation networks. We develop a multi-criteria resilience assessment framework and an optimized recovery strategy using a Genetic Algorithm (GA) that dynamically balances topological connectivity and functional demand. Validation on Beijing’s large-scale network shows our GA-driven strategy (RS4) significantly outperforms benchmark methods—random, topology-based, and function-based recovery—across disruption levels of 30% to 70%. The performance gap widens with disruption severity, demonstrating RS4’s robust efficiency in extreme scenarios and offering a practical tool for enhancing urban transport resilience.