Railway systems are now facing an increasing number of threats such as aging infrastructures and climate changes. The identification of critical network sections provides infrastructure managers with the ability to understand the impact of a disruption and creates a suitable preventive strategy to counter such threats. To this end, various vulnerability analysis methods have been proposed for railway networks. Two main types of methods, network topological analysis and network flow-based analysis, have been developed. Both approaches are constructed based on macroscopic models, which take only some railway properties such as network structure, train and passenger flow into account. Thus, the results obtained are high level approximations. This study proposes a new analysis method, which is developed based on the stochastic-microscopic railway network simulation model. The method can be applied to identify the critical sections of a railway network. The effect of impact levels and occurrence times of a disruption on the network section criticality is presented. An application of the proposed model is demonstrated using the Liverpool railway network in the UK.
Discerning Primary and Secondary Delays in Railway Networks using Explainable AI
