A Complex Heterogeneous Network-based Analysis Approach for Exploring Railway Operational Accidents

Abstract

Railway operational accidents are caused by a variety of related hazards’ interactions, which can be shown in the form of a complex network. By analyzing the topological characteristics of such a network, we can understand the causes of the accidents more deeply and make effective countermeasures. In this paper, a new analysis approach to understanding railway operational accidents based on a complex heterogeneous network is proposed. Its originality is to analyze accidents through topological indicators which are more suitable for heterogeneous networks and to design the targeted analysis process based on these indicators. Besides, the causality between other hazards or accidents directly related to one hazard in the network is quantified, which can describe the causal relationship in the network more accurately. This analysis approach provides a decision-making basis for ensuring the safety of railway operations. This paper takes the British railway operational accidents as a case study. The results show that this approach can effectively determine the key factors affecting the accidents and give a feasible decision-making basis.