Recognizing Critical Stations in Urban Rail Transit Networks Based on the PCA-TPE Method: Shanghai Metro as an Example

Abstract

The identification of core stations in urban rail transit (URT) networks remains a vital issue in network structure organization analysis and an integral part of network reliability evaluation. However, the identification of critical stations with a single centrality metric has limitations and the varying interactions between stations cannot be ignored. In this paper, a novel integrated approach is proposed by using the principal component analysis and topological potential considering entropy (PCA-TPE) method. Taking the Shanghai metro (SHM) network as a case study, a Space L network model is constructed and the network topology characteristics are analyzed. Moreover, the susceptible-infected (SI) model and the network failure simulation are employed to demonstrate the effectiveness of the proposed method. The results show that the SHM network exhibits characteristics of both small-world networks and scale-free networks. According to the experiments of the SI model, the nodes obtained by the PCA-TPE method have stronger spreading influence than those derived by other methods, especially in the initial stage. The failure simulations illustrate that attacks against the nodes detected by the PCA-TPE method will lead to devastating network failures. Hence, the proposed method is effective for identifying critical nodes in URT networks, and the findings of the research can provide theoretical evidence for the development planning and emergency management of the public traffic system.