Demand-based capacity assessment using mixed integer programming

Abstract

Understanding railway network capacity and potential reserves proves crucial for optimal planning decisions, which are required due to the high utilization of European railway networks and the intended future modal shift to rail. Capacity reserves emerge where the network structure does not match the demand. We propose a mixed integer program (MIP) for investigating the demand-based capacity reserves in the network. Extending the MIP-based railway network utilization model by adding demand structures, our model allows to optimize the train ordering and the locations of trains in the network, whilst regarding for the traffic demand to be served. The infrastructure is mesoscopically modelled by the individual blocks of the railway system. The demand is represented by a corresponding line plan and its given frequencies. The model determines the interrelations of demand and network capacity and thus allows to investigate between transport demand and network infrastructure. We test the proposed model on different artificial networks and a case study. In particular, the results show clear capacity effects of mismatched infrastructural demand and supply. It is thereby shown, that the efficient use of network capacity depends on the fit between demand and network structure. Furthermore, we can see that the emergent utilization behaviour is network specific and often non-linear, which strengthens the necessity of network approaches for global capacity assessment. Also providing support to other fields such as urban planning, the models incorporation to integrated and interdisciplinary planning approaches is left for future research.