Importance Analysis of Xi'an Metro Stations Based on K-shell Decomposition and Passenger Flow

Urban rail transit plays an important role in urban systems. Key stations in rail transit always transfer many passengers in a short time, which depends on the transition ability of the urban rail transit network. So it is essential to identify key stations for enhancing the transit system reliability. However, network topology is used in the reported research work to identify the key nodes in transit networks without consideration of attributes for stations. This paper integrates the structure features based on complex network theory and attributes features based on passenger flow to identify the key stations. The K-Shell-Passenger flow (KSP) method is developed by combining the K-shell decomposition algorithm with passenger flow information. Taking Xi'an metro as an example, Degree Centrality (DC), classical K-shell decomposition method (KS), and Degree centrality & K-shell decomposition method (DC+KS) are introduced to analyze the importance rankings of stations for verifying the effectiveness of the proposed method. The mean Average Precision (mAP) is used to verify the effectiveness of the proposed method quantitatively. The mAP of KSP is 99.67%, which is 7.4% higher than that of DC+KS method. Therefore, KSP is effective to analyze the key stations of the transit network, which can provide a targeted preventive maintenance strategy for stations based on the importance measures for improving operational efficiency.