ADVANCED MODELING AND CONTROL OF INTERMODAL TERMINALS AND RAILWAY NETWORKS

This PhD thesis presents the results of research on Information and Communication Technologies (ICTs) in the context of Smart Cities, with particular attention to the study, design, and the development of advanced models and control techniques for intermodal freight transport terminals and railway transport networks. At the moment, the implementation of a Smart City environment is largely recognized as an effective manner to improve the quality of life in the urban context. This thesis mainly focuses on the improvement of intermodal and railway transport, which are leading alternatives to road transport for the reduction of greenhouse gas emissions. In particular, the aim is to strongly increase the benefits that such transportation systems can generate by contributing to the resolution of the respective main managerial challenges. In particular, the intrinsic discrete event dynamics of (1) intermodal freight transport terminals and (2) railway transport is here considered to derive advanced models and control techniques to be used for the resolution of the main strategic/tactical and operational decision problems characterizing such systems. First, the viability of discrete event methods for smart transportation systems is here discussed. On one hand, it is provided a review of contributions on Petri nets for freight transportation systems; on the other hand, an overview on the discrete event MILP models for the railway rescheduling problem. On the basis of the developed reviews, contributions are here provided on (1) modeling, simulation, analysis, and control via Petri nets of Intermodal Freight Transport Terminals (IFTTs) and (2) on discrete event MILP modeling of railway traffic and the corresponding smart management when unexpected events occur in the network. Regarding topic (1), first a general modelling framework based on timed PNs is proposed which allows simulating and evaluating the performance of such key elements of the intermodal transportation chain. Then, it is shown how first-order hybrid Petri nets can be efficiently used to model and subsequently manage intermodal freight transport terminals by optimizing the terminal performance under alternative control policies Finally, it is demonstrated how timed Petri nets and the Data Envelopment Analysis (DEA) multi-objective optimization technique can be combined for the planning of intermodal terminals. The effectiveness of all of these techniques is tested on a real case study showing their practical use and ease of application. Regarding topic (2), first, a Decision Support System (DSS) for real-time management of railway networks is presented, which employs a MILP approach addressing traffic rescheduling under unexpected disturbances in a mixed- (single- and double-) tracked network. Then, it is proposed a self-learning decision making procedure for robust real-time train rescheduling in case of disturbances. The procedure is applicable to aperiodic timetables of mixed-tracked networks and it consists of three steps. The railway service provider can take advantage of this procedure to automate, optimize, and expedite the rescheduling process. Moreover, thanks to the self-learning capability of the procedure, the quality of the rescheduling is improved at each reapplication of the method. Both the presented techniques are applied to a real data set to test its effectiveness. Finally, the last contribution presents an innovative bi-level algorithm aiming at finding a feasible timetable for a mesoscopic rescheduling problem in case of disruption in a short computation time. It consists in the sequential resolution of two optimization MILP problems. It is here preliminary demonstrated, by the application of the method to a real case study (i.e., the national Dutch railway network), that the bi-level solving algorithm can be suitable for a real-time control environment thanks to its short computation time.