TTPlib 2008 – A Library For Train Timetabling Problems
Free (open access)
605 - 614
B. Erol, M.Klemenz, T. Schlechte, S. Schultz & A. Tanner
We introduce (TTPlib), a data library for train timetabling problems that can be accessed at http://ttplib.zib.de. In version 1.0, the library contains data related to more than 100 scenarios. Most instances result from the combination of macroscopic railway networks and several train request sets for the German long distance area containing Hanover, Kassel and Fulda, short denoted by HA-KA-FU. In this paper, we introduce the data concepts of TTPlib, describe the scenarios included in the library and provide a free visualization tool TraVis. Keywords: train timetabling problem, data library, optimization. 1 Introduction Mathematical optimization has greatly advanced many planning procedures in railway industry, like public transport network design, vehicle routing and dispatching or crew scheduling – however, the problem of computing new timetables from scratch is still unsolved. It is computationally not too difficult to verify that a given timetable is implementable, or to extend a given timetable incrementally by additional trains, but the computation of a timetable that is optimal in some mathematically well-defined sense is provably hard. For similar problems of mathematical optimization, the establishment of standard models and standard problem libraries have contributed to the success in problem solving. Such libraries exists for the famous Traveling Salesman Problem,
train timetabling problem, data library, optimization.