Integrated timetabling, vehicle scheduling, and dynamic capacity allocation of modular autonomous vehicles under demand uncertainty (2410.16409v1)

Abstract: The Integrated Timetabling and Vehicle Scheduling (TTVS) problem has extensive applications in all sorts of transit networks. Recently, the emerging modular autonomous vehicles composed of modular autonomous units have made it possible to dynamically adjust on-board capacity to better match space-time imbalanced passenger flows. In this paper, we introduce an integrated framework for the TTVS problem within a dynamically capacitated and modularized bus network, taking the time-varying and uncertain passenger demand patterns into account. The fixed-line modularized bus network operates units that can be (de)coupled and rerouted across different lines within the network at various times and locations to respond to the time-varying demand, providing passengers with the opportunity to make in-vehicle transfers. We formulate a stochastic programming model to jointly determine the optimal robust timetable, dynamic formations of vehicles, and cross-line circulations of these units, aiming to minimize the weighted sum of operational and passengers' costs. To obtain high-quality solutions of realistic instances, we propose a tailored integer L-shaped method coupled with valid inequalities to solve the stochastic mixed-integer programming model dynamically through a rolling-horizon optimization algorithm. An extensive computational study based on the real-world data of the Beijing bus network shows the effectiveness of the proposed approaches. Our method outperforms the two-step optimization method involving sequential decision-making for timetables and vehicle schedules. Furthermore, the computational results illustrate that our approaches are able to find timetables and vehicle schedules requiring fewer units and lower operational costs compared with using fixed-formation vehicles.