Optimization of Turn-Back Station Locations in Urban Rail Transit Networks Considering Disruption Uncertainty

Urban rail networks play a crucial role in public transit systems supporting daily commuting in metropolitan areas. Turn-back stations, which allow trains to reverse direction, are essential for maintaining the system’s level of service during disruptive events. Optimizing their locations can enhance the network’s resilience to disruptive events. Considering the uncertainty of failures in the postevent network, this study develops a stochastic optimization model for optimizing the location of turn-back stations to improve network resilience. A resilience metric is proposed that accounts for both the number of passengers reaching their destinations and the total travel time within the system. To optimize the locations of turn-back stations, various failure scenario sampling techniques are employed, either to enhance the mean network performance across the sampled scenarios or to reduce the variance of performance. A case study of a simplified Beijing Rail Transit network demonstrates that the proposed location optimization model significantly improves the system resilience compared to existing solutions. Additionally, a sensitivity analysis of the number of turn-back stations is conducted. The research findings provide effective decision support for the design of resilient urban rail networks against disruptive events with uncertainties.