MAS-based Model for Evaluating Train Timetables to Minimise the Waiting Time

In recent years, continuous developments in urban rail transit have led to automatic fare collection systems being installed in various cities. This not only improves passenger travel efficiency but also allows information on the behaviour of passengers in the system to be collected and used to optimise the train timetable. This paper presents an optimisation model for determining the optimal headway for a train timetable. In addition, a Multi-Agent System (MAS)-based model is presented for simulating the interactions between passengers and trains to estimate the locations of passengers in a rail network at a given time. The MAS-based model was applied to simulating the actual operation and predicted long-term demand of a newly built metro line to validate its applicability to urban rail transit networks, and the results were used to determine the optimal headway for solving the mismatch between the transport capacity according to the timetable and demand according to the passenger flow volume. The simulation results can be used as a data basis for the design and adjustment of train operating plans.