Bayesian network-based risk evaluation model for the operational requirements of the China Railway Express under the Belt and Road initiative

Abstract

The operation of the China Railway Express features numerous links across several regions and countries. Effectively controlling the risks involved in the operation of the China Railway Express is crucial for ensuring safety and efficiency and promoting the sustainable development of the China Railway Express. The Bayesian network-based risk-management model was built corresponding to the actual operation of the China Railway Express; the model used an advanced risk-management theory. The sensitivity analysis of risk factors and Bayesian network inference were realized using the expectation-maximization and clique-tree propagation algorithms. Using the risk-checklist method, 17 risk-related factors were analysed on 17 nodes of the Bayesian network from three perspectives—safety, efficiency and effectiveness—based on expert opinions and the actual operating conditions of the China Railway Express. Data from a sensitivity analysis and evidence inference of the Bayesian network model indicated that the sensitivity coefficients of nodes N01, N04, N07, N08 and N17 of the network were high. Moreover, the risk-occurrence probabilities for nodes N01, N04, N06, N07 and N09 were higher in the case of reverse inference. Our results revealed the crucial factors influencing the risk. The identified risk factors included the stability of the political environment in countries along the route, conditions of station infrastructures and the complexity of the process of changing rails and reloading. Further, risk-management suggestions were provided. By establishing a sound risk-management framework, reliable assessment and management could be realized in accordance with changes in the operating conditions of the China Railway Express.