Intermodal Transfer Coordination in Logistic Networks

Thumbnail Image


Publication or External Link






Increasing awareness that globalization and information technology affect the patterns of transport and logistic activities has increased interest in the integration of intermodal transport resources. There are many significant advantages provided by integration of multiple transport schedules, such as: (1) Eliminating direct routes connecting all origin-destinations pairs and concentrating cargos on major routes; (2) improving the utilization of existing transportation infrastructure; (3) reducing the requirements for warehouses and storage areas due to poor connections, and (4) reducing other impacts including traffic congestion, fuel consumption and emissions.

This dissertation examines a series of optimization problems for transfer coordination in intermodal and intra-modal logistic networks. The first optimization model is developed for coordinating vehicle schedules and cargo transfers at freight terminals, in order to improve system operational efficiency. A mixed integer nonlinear programming problem (MINLP) within the studied multi-mode, multi-hub, and multi-commodity network is formulated and solved by using sequential quadratic programming (SQP), genetic algorithms (GA) and a hybrid GA-SQP heuristic algorithm. This is done primarily by optimizing service frequencies and slack times for system coordination, while also considering loading and unloading, storage and cargo processing operations at the transfer terminals. Through a series of case studies, the model has shown its ability to optimize service frequencies (or headways) and slack times based on given input information.

The second model is developed for countering schedule disruptions within intermodal freight systems operating in time-dependent, stochastic and dynamic environments. When routine disruptions occur (e.g. traffic congestion, vehicle failures or demand fluctuations) in pre-planned intermodal timed-transfer systems, the proposed dispatching control method determines through an optimization process whether each ready outbound vehicle should be dispatched immediately or held waiting for some late incoming vehicles with connecting freight. An additional sub-model is developed to deal with the freight left over due to missed transfers.

During the phases of disruption responses, alleviations and management, the proposed real-time control model may also consider the propagation of delays at further downstream terminals. For attenuating delay propagations, an integrated dispatching control model and an analysis of sensitivity to slack times are presented.