Modelling the Traffic Prediction at the Direct Level of TMS Systems

The paper deals with the issue of modelling traffic prediction in operational applications of infrastructure managers to increase the quality of decision-making at the direct level of railway traffic management. First, the paper defines traffic prediction and the influences that can generate traffic conflicts. Then, it proposes an approach to modelling traffic prediction that eliminates the generated traffic conflicts and evaluates the benefits of the proposed model. The authors’ proposal is based on their knowledge of the direct-level Traffic Management System (TMS) in the Czech Republic. The aim is to propose options for traffic prediction solutions to be used primarily on the railway network of the Czech Republic, but also elsewhere. The solutions are proposed with regard to their realizability in currently used computing resources.