The operation of a mass rapid transit (MRT) should consider the balance between total system costs and service level. For a MRT system, the main service is to provide passengers with the cost-affordable mobility, under the regulation of MRT service indicators; the government sets the minimum standards to ensure a certain level of MTR services. Thereby, how to establish an optimal operating timetable is one of the important operational issues for a MRT system.
In the past research concerning MRT operations, most researches focused on the issues of train delay, energy saving, route design or overall system operating regards, and placed less focus on the optimal timetable problem. In the Kaohsiung Mass Rapid Transit (KMRT) system, the total ridership has not reached a predicted level, but the service provided in terms of the Train Service Plan (TSP) is more than needed. Therefore, a time-dependent time table for the KMRT system is crucial to minimize total system cost while maintaining the certain level of train service.
In this research, we aim to develop a time-dependent timetable model which is able to dynamically adjust train schedule depending on the passenger spatio-temporal distribution demands during daily operation. The developed model will be solved by minimizing both the operating cost and passenger waiting cost. Finally, numerical case study and sensitivity analysis will be conducted to demonstrate the feasibility and effectiveness of the developed models and solution algorithms.
Mass Rapid Transit System Passenger Traffic Forecast Using a Re-Sample Recurrent Neural Network
