Author(s)

J. Rodriguez

Abstract

Empirical study of a railway traffic management constraint programming model J, Rodriguez Institut National de Recherche sur les Transports et leur Securite, 20, rue Elisee Reclus, F 59650 Villeneuve d'Ascq, France Abstract INRETS and SNCF are involved in a project to develop a decision support system for traffic control centres. A specific goal of the project is to develop different models of train circulation management problems and experiment heuristics to solve them. In a previous paper[10], the train scheduling problem was formulated as a joint scheduling and allocation problem, i.e. there are several degrees of freedom for deciding both which activities to perform and when, and which resources to allocate to these activities. In this paper we will present some experimental results of the model. Initial experiments showed weak time performances. Afterwards, with the improvement of constraint propagation, the choice of better heuristic commitment techniques and a t

Keywords