The timetabling step in railway planning is based on the estimation of the running times. Usually, they are estimated as the shortest journey duration increased of a short time supplement. Estimating the running time amounts to define the speed profile which indicates the speed that the train driver must hold at each position. The approach proposed in this paper produces a set of solutions optimizing both the running time and energy consumption. The approach is based on an original method of speed profiling performed by a multi-objective evolutionary algorithm. The speed profiles found by the evolutionary algorithm are all compromises between journey duration on the one hand and energy consumption on the other hand. A set of results obtained on two real-life instances are analyzed and discussed to highlight the relevance of such an approach in an operational context.
