Solution methods for pricing and fleet management in shared automated vehicle services considering supply-demand dynamics, congestion, and income heterogeneity
Qiaochu Fan (TU Delft - Civil Engineering & Geosciences)
J. Theresia van Essen (TU Delft - Electrical Engineering, Mathematics and Computer Science)
Gonçalo Homem de Almeida Correia (Universidade Técnica de Lisboa)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Research on the supply–demand dynamics of Shared Automated Vehicle (SAV) services has grown rapidly. However, the resulting models are often highly complex nonlinear systems that are difficult to solve. This study has two main objectives: (1) to investigate profit-maximising pricing and fleet management strategies for SAV services while accounting for supply–demand interactions, congestion effects, and travellers’ heterogeneous income levels; and (2) to develop and compare several solution methods for determining these strategies. We formulate a mixed-integer nonlinear programming model for the SAV fleet management problem under three pricing strategies: a base fare plus a distance-based fare, a distance-based fare only, and an income class–based fare. To solve this model, we propose three algorithms designed to address its nonlinearities: a linearisation-based method (P1), a hybrid particle swarm optimisation method (P2), and a hybrid Bayesian optimisation method (P3). We also examine the feasibility of the solutions obtained by these methods with respect to the original problem, and introduce a correction procedure when infeasibility arises. The proposed methods are evaluated through a small case study and a larger case study based on Delft, the Netherlands. The results show that all three methods can find good-quality solutions within short computational times in the small case study, while P3 achieves the best objective value in the Delft case study. From a practical perspective, the Delft case study shows that the first pricing strategy can encourage active transport for short trips, the second leads to the lowest congestion but also the lowest profit, and the third achieves the highest profit by attracting more demand and exploiting the lower price sensitivity of high-income travellers, despite causing higher congestion. Additional analyses further examine the effects of income class segmentation, income distribution, and congestion.