Modelling marshaling yard processes using queueing theory

Abstract

Abate Sewagegn. Modelling the marshalling yard processes using queueing theory. Ostrava 2025. Dissertation. VSB-Technical University of Ostrava. Supervisor doc. Ing. Michal Dorda, Ph.D. Marshalling yard has been an essential part of freight rail transport operations by dispatching and organising loads based on their destination. However, due to the coupling and decoupling time consumption and capacity old yard results in longer dwelling time and delay in transportation. The dissertation investigates the modelling and performance of the marshalling yard process using queueing theory. The study implements analytical methodology by developing M/M/1/m, M/Ek,k-1/1/m, M/HypoK/1/m, Er/Es/1/m, and M/C2/1/m queuing models that incorporate service line failure. We employed analytical models, validated using CPN simulation models, to evaluate system performance. Based on the integrated models, key performance measures, including the expected number of customers in the service, queue, system, and failed service line, were calculated, and sensitivity analysis was tested. The output of models indicated that the marshalling yard can be effectively modelled. This research contributes to the scientific discipline by extending classical queuing models to capture multi-phase dynamics and sensitivity effects, and by providing practical guidance for optimising yard operations. The study also identifies directions for future research, including the exploration of more complex phase-type distributions and real-time optimisation.