Metro timetable optimization from passenger perspective based on simulation models and incomplete data of passenger flow

Due to the structure of metro railway systems, most of its planning problems including timetabling are classified as optimization problems. Timetabling is the problem of assigning times for events such as departure and arrival of trains over the planning time horizon. As modern metro systems try to provide customer centric services, it is required that the timetables be optimized to minimize the passenger waiting times, and avoid high passenger loads inside the trains. Introducing passenger demand adds significant complexity to the problem. This paper adopts a simulation approach to achieve such an optimization. The main idea is to iteratively generate feasible timetables, and evaluate them by means of simulation. The simulation includes both the train operations and the passenger flow. A real world use case is also demonstrated.