Redundancy is a crucial component of network resilience. To address the current gap in time redundancy consideration when identifying critical stations in rail transit networks, this study proposes a comprehensive approach. First, we develop a timeweighted rail transit network model incorporating Space L and multiline coupling, which accounts for interstation transfer times. Second, we employ node deletion to simulate station disruptions and introduce three metrics: station redundancy index, network redundancy index, and a redundancy evaluation model based on shortest travel time variations. Using the 2024 Chongqing rail transit network as a case study, we compare the results between redundancybased and betweennessbased evaluations. The analysis reveals an average station redundancy index of 0.7998, with 3.87% of stations showing a Kmeans clustering center of 0.2788, indicating lower redundancy at critical locations. Redundancycritical stations are predominantly located near the network periphery along extended lines, while betweennesscritical stations are typically found at multiline interchange points, demonstrating distinct spatial patterns between these two categories. Network performance is more significantly impacted by disruptions at redundancycritical stations compared to betweennesscritical stations, validating the effectiveness of our identification method. This approach provides valuable insights for enhancing the resilience of urban rail transit networks.