Study on the Evolution and Resilience of Rail Transit Time Networks—Evidence from China

In the network operation and management of rail transit systems, the occurrence of unexpected events causes damage to the network structure, further hindering regional accessibility performance and the function of the system. This study is based on the rail transit operation schedules in 2009, 2013, 2016, 2019 and 2022. We construct a directional weighted rail transit time network (RNNT) with train operation time as the weight, compare the betweenness centrality, sum of the shortest time path and entropy importance, etc., and quantitatively measure the network accessibility, connectivity and its resilience evolution. The results show that the current rail transportation network in China has a “small-world” effect, and there are a few stations with strong connections. The most densely distributed intervals of travel times between pairs of nodes changes from 440, 445 to 207, 210. The fastest and best-performing disturbance to network connectivity and accessibility performance are both caused by the betweenness disturbance strategy. When the network connectivity remains 80% effective, the ratio of failed nodes under the static betweenness centrality strategy decreases from 3.96% in 2009 to 2.31% in 2022, with weaker connections between node pairs, and their network resilience diminishes. When the network accessibility remains 80% effective, the ratio of failed nodes under the static (dynamic) betweenness centrality strategy increases from 0.13% (0.13%) in 2009 to 0.20% (0.23%) in 2022. Therefore, the rail transit network can protect the corresponding rail stations based on the station ranking of the above strategies, and this research is beneficial to rail transit network protection and structure optimization.