Resilience assessment and recovery strategy optimization for bus-metro composite network under emergency scenarios

Urbanization has resulted in increasingly complex transportation networks, rendering the enhancement of disaster resilience in bus and metro systems a critical research priority. This study develops a topological-functional integrated framework for the Bus-Metro Composite Network (BMCN), overcoming limitations of conventional network analysis. An optimal recovery model is established using the Soft Actor-Critic (SAC) algorithm, with the objective of maximizing resilience. The study focuses on the BMCN within the rectangular area (approximately 3.8 km × 3.4 km) between Xuejiadao Station and Lingshanwei Station in the West Coast New Area of Qingdao City. A comparative analysis of four strategies under small-scale and large-scale failure scenarios demonstrates that the SAC-based recovery strategy outperforms the others in both failure scenarios. Furthermore, its effectiveness is more pronounced in large-scale failure scenario than in small-scale one. The findings provide decision-making support for transportation authorities in emergency management and contribute to advancing resilient urban development.