A heuristic approach to integrate train timetabling, platforming, and railway network maintenance scheduling decisions

Train timetabling, platforming, and network maintenance scheduling are three highly interdependent problems that are crucial in the planning of railway operations, and each is normally addressed separately. In this paper, we simultaneously optimize these problems for a highspeed railway network that is comprised of multiple railway lines and stations. We model the railway network on a mesoscopic level and formulate a 0-1 binary integer programming model that minimizes the total train weighted running cost and any deviation from ideal maintenance task start times. A heuristic procedure, which dynamically updates the available time windows for each of the trains, is used to control the number of train paths in the mathematical model. The mathematical model is repeatedly solved, and at each iteration we gradually modify the set of train paths available. Four different strategies to modify train time windows are used in the train path modification step and their selection depends on the solution to the mathematical model.