Weiting Zhao, Ullrich Martin, Yong Cui and Maureen Kösters
International Journal of Railway, vol. 9, no. 1, pp.1-9, 2016

Abstract : In order to achieve a comprehensive utilization of railway networks, it is necessary to accurately assess the timetableindicators that effect the train operation. This paper describes the parameter calibration for two timetable indicators:scheduled running time and scheduled dwell time. For the scheduled running time, an existing model is employed andthe single timetable parameter (percentage of minimum running time) in that model is optimized. For the scheduleddwell time, two intrinsic characteristics: the significance of stations and the average headway at each station are pro-posed firstly to form a new model, and the corresponding timetable parameters (the weight of the significance and theweight of the average headway) are calibrated subsequently. The Floyd Algorithm is used to obtain the connectivityamong stations, which represents the significance of the stations. A case study is conducted in a light rail transportationsystem with 17 underground stations. The results of this research show that the optimal value of the scheduled runningtime parameter can be automatically determined, and the proposed model for the scheduled dwell time works well with ahigh coefficient of determination and low relative root mean square error through the leave-one-out validation.

Keyword : Calibration, Scheduled Running Time, Scheduled Dwell Time, Floyd Algorithm, Simulation of Railway Opera-tion, Capacity Research