Improving an Evolutionary Multi-Objective Approach for Optimizing Railway Energy Consumption

AbstractIn this paper, an approach for speed tuning in railway management is proposed in order to optimize journey duration and energy consumption. Indeed, in railway transportation, fastness of travel has often been preferred to other criteria, such as energy. However, for recent years, the concern due to pollution and global warming leads to develop more eco-aware transportation systems, notably for the existing rolling stocks. Nevertheless, reducing energy consumption is, most of the time, achieved at the expense of journey durations.In order to obtain compromises between journey duration and energy saving, this approach deals with a bi-criteria optimization problem, which consists in designing speed profiles integrating both criteria in order to provide patterns of speed control. For that reason, the main objective of the proposed method is to obtain a set of solutions, diversified enough to help decision-makers to choose the most suited speed profile. Based on a state-of-the-art multi-objective evolutionary algorithm, the approach searches for diversified solutions in a continuous space. So as to evaluate these solutions, the Pareto approach is used to assess and rank them. Validating the method is done by applying it on a reallife case study: the Saint-Etienne–Rive de Giers line in France. The outcomes are then detailed and analyzed in order to highlight the efficiency of the method