SYNCHRONIZATION OF THE LAST-TRAIN TIMETABLE CONSIDERING PASSENGER TRANSFER TIME

Unsynchronized integrated lines in urban rail systems may cause some passengers to miss their last transfer chance. Therefore, passengers move faster in a swarm mentality to catch the last train. This study aims to synchronize the last-train timetable so that the maximum possible number of passengers can be transferred at the transfer stations by waiting for the minimum time. This paper provides a new approach to show that passengers who try to get to the last train at the transfer station move faster and describes the transfer processes with heuristic algorithms. In the case of the Istanbul urban rail system, passengers transfer to the last train 32% faster than average. To find the heuristic algorithm that defines the transfer processes of these passengers, particle swarm algorithms, dragonfly algorithms, and a simulated annealing algorithm were selected for comparison. The transfer times obtained with the particle swarm algorithm and the actual transfer times gave close results between 97.6% and 99.2%. The modified last-train timetable with predicted transfer time increases the number of successful transfers by 28%, decreasing the average waiting time of passengers from 197.27 seconds to 50.56 seconds. In addition, passengers wait 58 seconds less for the transfer to the last train by adjusting the timetable to the modified last train transfer time.