STOCHASTIC MODELING OF DELAY PROPAGATION IN TRANSPORTATION SYSTEMS FOR CAPACITY OPTIMIZATION
DOI:
https://doi.org/10.26577/JMMCS131320269Keywords:
virtual coupling, freight trains, minimal headway, stochastic disturbances, timetable reliability, discrete-event simulationAbstract
The minimal headway for virtually coupled (VC) freight trains operating under stochastic timetable disturbances on the Trans-Siberian Railway is investigated. By analyzing empirical data from the Oy-Mos section, we identify distinct delay distribution patterns for regular freight (Erlang), heavy-haul (chi-square) and extra-long trains (Weibull). Using a hybrid discrete-event simulation (DES) and Petri net model, the impact of reduced headways (up to 3 min) on the capacity and timetable reliability is evaluated. The results reveal that headways below 6 min trigger secondary delay propagation, while 6-minute intervals are optimal for the capacity without compromising robustness. The adjustment of the interval from 13 min in the current automated blocking system to 7 min allows for a 54.2% increase of capacity for regular freight trains, 53.3% for heavy-haul trains, and 58.6% for extra-long ones. The proposed framework integrates real-world disturbances into the VC planning, offering a novel approach to the freight railway optimization.











