Strategic Rail Network Optimization Model for Freight Transportation

Abstract

An optimization model for a rail freight network was developed as a support tool for planning and policy decisions involved in the improvement of rail networks on a regional and national level. The model was based on a strategic traffic assignment model, which was designed to model macronetworks with a high aggregation level, and was exclusively designed for freight traffic. The model contemplates road and rail transport modes and considers two types of cargo: intermodal cargo, which is generally transported in containers and is easily interchanged between different modes at intermodal terminals, and general cargo, which represents all the remaining cargo. The optimization process was based on a local search heuristic that delivers good solutions in a reasonable computing time. The quality of each network improvement solution was assessed on the basis of the reductions in total generalized costs and carbon dioxide emissions. This optimization model for a freight network is innovative because it is not limited: both the improvement of existing links and the construction of new ones are allowed, and the model does not have a limit on the number or variety of possibilities for network improvement. The adaptability to different conditions was emphasized when the model was applied to two networks under different investment scenarios because the model delivered considerably different solutions adapted to the conditions of each scenario.