Model predictive control for a sustainable transport modal split at intermodal container hubs

The increase of international commerce and the expected container vessels capacity with 18, 000 TEU (twenty-foot equivalent unit) will put a considerable pressure on container hubs. High flow peaks will appear at gateway hubs in the transport network compromising the cargo transportation towards the hinterland and decreasing the network transport capacity. Moreover, authorities are forcing transport operators to operate in more sustainable ways. For container hubs this is translated into making a preferable choice for barge and train modalities before opting for truck modality. In this work we present a framework based on Model Predictive Control (MPC) to address the so-called transport modal split problem for the outgoing cargo at container hubs. We use two features (destination and due time) to categorize the cargo present at a container hub and develop a dynamic model to make predictions of cargo volume over time. The controller decision takes into account transporting cargo towards the final destination while opting for sustainable transport modalities. The approach is able to assign cargo in advance to the existing connections at the hub in order to overcome predicted cargo peaks in the future. The framework can also be used to choose between different connection schedules. Giving decision freedom to container hubs is a step towards a synchromodal and more flexible transport network. These statements are illustrated with two simulation examples.