Abstract:

As a typical representative of the new mode of shared public transport, demand responsive transit is facing the challenge of efficiently processing travel demand and real-time planning of vehicle routes. Traditional research on demand responsive transit dynamic scheduling mainly focuses on real-time adjustment of vehicle routes after dynamic demand is known, which often struggles to comprehensively adapt to changes in travel demands. Therefore, this study introduces model predictive control methods to establish a dynamic scheduling model for demand responsive transit, utilizing potential future-stage passenger flow information to optimize current-stage scheduling decisions and timely re-planning according to the latest disclosed information. In this study, an adaptive large neighborhood search strategy is adopted to optimize the iterative vehicle scheduling sequence through a two-stage heuristic method. Numerical experimental results demonstrate that in ideal scenarios without prediction deviation, compared to traditional dynamic scheduling methods, this study's approach significantly reduces the total cost of the system by 14.54%; even in a pessimistic scenario with a 30% prediction deviation, it still achieves a cost optimization of 5.27%. Moreover, various passenger service indicators exhibit superior performance, indicating strong universal applicability in different stochastic environments.

Key words: transportation engineering, demand responsive transit, dynamic scheduling, model predictive control, adaptive large neighborhood search algorithm