LAUSR

In this paper, the possibility and potential benefits of implementing discriminatory policies in rail freight transportation are analyzed, with the aim of revenue maximization. A regular, cyclic, single train service with fixed composition and capacity is studied. The problem is decomposed into discrete time periods. Transportation requests arise randomly over time, and the decision of either accept or reject a certain request has to be made. The problem is formulated via dynamic programming, and the deterministic approximations of the problem are used in order to formulate booking limits and bid price policies. Results obtained are compared with those of standard first come – first served policy, which is implemented by the Serbian railways. Although not acceptable in all contexts, the proposed aggressive policies demonstrated promising benefits.