Using AIE-D algorithm to recognize the node importance of weighted urban rail transit network considering passenger flow

Abstract

The AIE-D algorithm (Adjacent Information Entropy-D algorithm) is proposed to recognize the importance of nodes in the urban rail transit network (URTN) weighted by passenger flow, which considers passenger flow, topological characteristics of nodes in the URTN, and the influence of neighboring nodes. The travel impedance is determined by using travel time, the D algorithm is used to search the k-short paths, and the weight value of each edge is the passenger flow cross-section of the corresponding line. Then, the detail AIE calculation steps are introduced. Next, a numerical study and comparison study are conducted by using the weighted topology of network. Compared with other commonly used algorithms, AIE-D has lower time complexity with faster calculation speed, and higher recognition accuracy. Finally, a real-world case study is conducted by using URTN of Chengdu Metro Network as the background. Weighted by passenger flow has greater impact on the operation of urban rail transit. The nodes are categorized into three classes according to the ranking of node importance, which includes Classification VI, Classification I and Classification GI. We conduct random attacks and deliberate attacks on the network, and analyze the network efficiency and maximum connectivity subgraph rate after the attacks.