Resumen
Determining the interventions, e.g., maintenance, renewal, improvement and extension, to be included in an infrastructure program requires the consideration of the asset, intervention, traffic, and network characteristics. This, in turn, requires the development of an appropriate system model enabling the construction of straightforward optimisation models. Although there are already a considerable number of such system models in the literature, improved modelling of the complex relationships between interventions, intervention costs and the service provided by the infrastructure network is possible?especially in the trade-off between the accuracy of considering the complex relationships and the simplicity of the mathematical formulation. This paper explains how to build system models for railway infrastructure networks that capture the complex relationships in a system model that can then be used to construct mixed integer linear optimisation models. The proposed type of system model includes how both intervention costs and impacts on service vary as a function of the type, time and location of the interventions included in intervention programs. The system models of this type consist of a graph that is used to model the relationship between the interventions and intervention costs on the asset level, and the relationship between the interventions and the service provided on the network level. The algorithm uses systematic intervention classification and a hierarchical network state structure to build the system model. For illustration purposes, a system model for a railway network consisting of five track segments, seven switches, a bridge, a tunnel and the power supply system is developed using the algorithm.