Risk propagation analysis of urban rail transit based on network model

Abstract

Drawing on theories of complex network and accident causality, this paper explores the generation and propagation of urban rail transit (URT) risks, and predicts the propagation path and law of URT risks, aiming to prevent and control operational accidents. Specifically, the URT risks were defined and classified based on the records of accidents and faults. The risk analysis network for the URT was constructed in the light of global safety behavior. Then, state transition measure and coupling measure for risk nodes were proposed. Inspired by the classic dynamics model of infectious disease, the propagation and evolution of risks in the URT were simulated, and the risk link group with high propagation risks was deduced by using the two proposed measures. The proposed risk network was verified through MATLAB simulations on the bogie system of a train and the related risk nodes. The risk link group with high probabilities of risk propagation was successfully derived for the bogie system, an evidence of the feasibility of our dynamics model for URT risks. The research findings lay the basis for risk management and control of the URT systems.