Optimal intervention programs for railway infrastructures

Project description

Railway infrastructure managers face the task of planning interventions to ensure an adequate level of service. Thus, they develop intervention strategies defining how to manage the infrastructure in the long-term and determine intervention programs containing the interventions to be executed in the upcoming planning period. When determining intervention programs, infrastructure managers need to consider the interdependencies between the individual infrastructure assets. The network characteristic of the infrastructure brings possible benefits and constraints when interventions are grouped together.

This research projects aims to develop a methodology to determine optimal intervention programs for railway infrastructures. This includes a method for the quantification of the net benefit of an intervention program and an optimisation model that enables the determination of optimal intervention programs for railway infrastructures. The method for the quantification of the net benefit aligns the determination of intervention programs with the development of intervention strategies. It considers the short-term impacts related to the execution of the interventions, e.g. intervention costs and service disruption, and the long-term impacts related to changes in the risk and costs for future preventive intervention based on the improvement obtained by the interventions. The network flow optimisation model developed uses a graph-based system model that enables the modelling of economical and topological dependencies between interventions. This allows the consideration of synergies in the intervention costs and in the effect on the service provided when grouping interventions in an intervention program. The optimisation model further considers structural and resource dependencies that constrain the intervention selection. The graph-based structure allows the formulation of efficient mixed integer linear programs. The methodology is applied on a variety of realistic example networks to demonstrate the functionality of the optimisation model and its usefulness for infrastructure managers when integrated in a digital infrastructure management.