Demand-responsive transit (DRT) is a flexible public transportation mode offering affordable door-to-door services. However, its widespread adoption still faces large hurdles such as demand variability, immediacy, and financial sustainability. Most DRT studies focus on fleet management, often leading to underutilization of capacity due to passenger spatial dispersion. This issue calls for multi-objective optimization for both service coverage and cost efficiency. This study proposes a dynamic DRT scheduling problem that integrates vehicle-passenger coordination and time-dependent travel times, optimizing fleet management by leveraging passengers’ spatial and temporal flexibility. We propose a multi-objective optimization model within a rolling horizon framework to optimize vehicle routing, departure times, and passenger assignment, with dual objectives of maximizing profit and minimizing passenger spatiotemporal displacement. To solve this problem efficiently, we develop a dynamic multi-objective Memetic algorithm entailing three salient features: 1) distinguishing static and dynamic phases while identifying similar environments by comparing the scheduling records in the environment and the updated request pool; 2) using memory-based environment inheritance to accelerate multi-period decision-making; 3) developing a heterogeneous elite selection strategy during iterations to address the issues of speeding proliferation in dynamic problems. Our approach is validated through a real-world case study in Nansha District, Guangzhou, China. Results show that our algorithm performs comparably to benchmark algorithms in both solution quality and efficiency, and outperforms advanced methods across multiple metrics. Managerial insights are also provided.
https://doi.org/10.1016/j.tre.2025.104232Cite as:
@article{wu2025dynamic,
title={Dynamic demand-responsive transit scheduling with time-dependent travel times: A joint supply and demand management approach},
author={Wu, Weitiao and Zhang, Zeyue and Lu, Kai and Ren, Jingxuan},
journal={Transportation Research Part E: Logistics and Transportation Review},
volume={202},
pages={104232},
year={2025},
publisher={Elsevier}
}