The Identification of Human Safety Risks in Railway Train Dispatching Based on the DEMATEL-IANP Model

In large-scale, high-density railway networks, the safety of train dispatching is directly linked to transportation efficiency and the protection of human life. Accurately identifying and scientifically evaluating associated risks has become a critical issue in the management of railway operational safety. This paper proposes an integrated risk identification model based on the Decision-Making Trial and Evaluation Laboratory and an Improved Analytic Network Process (DEMATEL-IANP). Focusing on human-related factors in train dispatching, a comprehensive safety assessment framework comprising 13 key risk elements is established. Subsequently, the DEMATEL method, combined with a Backpropagation Neural Network (BPNN) optimized ANP approach, is applied to systematically analyze the interdependencies among these factors and to determine the relative weights of each indicator. Finally, the relative influence and significance of these factors within the overall railway safety system are evaluated. The results reveal that dispatcher fatigue and distraction, inadequate safety awareness, poor communication, and driver inattentiveness are the major human-related risk factors affecting dispatching safety. Notably, dispatcher fatigue and distraction exert the greatest impact, followed closely by insufficient safety awareness among associated personnel. As critical human risk factors in dispatching safety management, their effective identification and mitigation are essential for enhancing operational decision-making and strengthening overall railway safety control.