چكيده به لاتين
One of the most important problems in the rail transport industry is the problem of train scheduling. In the real world, scheduling parameters are often associated with uncertainty. In recent years, with the increasing complexity of rail systems and the need to use uncertainty planning techniques, researchers are increasingly focusing on uncertainty models and providing an efficient schedule because of an inadequate schedule, Will impose more travel time on passengers and will reduce their level of service and dissatisfaction. Among the uncertainty techniques, robust optimization is one of the new approaches to deal with uncertainty, which seeks to provide a solution to the problem that is based on data uncertainty. The basic idea in robust optimization is to consider the worst-case scenario and optimize based on the worst-case scenario. In this dissertation, the schedule of passenger trains along with the stop and transit program is studied under the conditions of uncertainty in demand. In the method of stopping and passing each train has a special stop pattern and each can pass a number of stations without stopping in order to reduce the total travel time of passengers and increase network performance. In order to find these patterns, in this dissertation, a definite model with the aim of minimizing the total travel time of passengers and an indefinite model with a stable style optimization approach with the aim of minimizing unmet passenger demands are presented. The limitations of the definitive model include the capacity of trains, the minimum guide, the limit of satisfying the demand of all passengers, the minimum stop time for each train at each station and the passenger tolerance threshold, and the limits of acceleration and braking times of trains. In order to consider the real world conditions, a definite model has been developed by considering the uncertainty in passenger demand. This has led to the presentation of a robust model with a robust style optimization approach. The results of solving the mathematical model are evaluated using a numerical example and a simulated sample of real data. In order to validate the robust model, the sensitivity of the existing parameters is analyzed and the logical behavior of the model is investigated.
