The resilience of an urban rail transit (URT) network when faced with disruptions is affected by the locations of stations equipped with turn-back (TB) tracks. However, limited studies have enhanced the resilience of a URT network by setting new TB tracks. The present work addresses this gap by proposing and solving a scenario model for improving the operation of a URT network under normal conditions and disruptions by considering uncertain disruptions. A solution algorithm combined with the non-dominated sorting genetic algorithm-II is proposed to solve the model. Numerical experiments conducted on the Chengdu subway system indicate that the resilience of a URT network is significantly affected by TB operations provided at stations equipped with TB tracks. Compared with a network without new TB tracks, the matching degree between passenger flow spatial distribution and TB convenience, and the network’s overall resilience metric (NORM) are improved by 12.05% and 0.58%, respectively, when five new TB tracks are installed. The solution effectiveness of the model is related to the number of new TB tracks, and the NORM decreases by an average of (Formula presented.) after adding new TB tracks to a station.