轨道网络加密和客流波动加剧使接驳需求呈现强实时性和时变性特征,传统模式在运力配置和服务响应上存在滞后问题。该文首先引入自动驾驶模块化车辆,提出模块化接驳公交(modular autonomous feeder bus,MAFB)多目标动态调度优化方法;其次,在建模层面,联合决策线路、发车时刻和车辆编组,构建了最小化乘客出行时间和运营成本的多目标优化模型;最后,在求解层面,基于滚动时域方法设计了多目标自适应大邻域搜索算法,并通过算子自适应和非支配解维护高效探索了动态可行域。结果表明:该文所提方法能在可接受时间内获得高质量Pareto解集;在乘客等待时间、绕行时间和运营成本等指标上均优于常规方案,实现了运营成本与服务质量的协同提升,验证了车容量和时域间隔对服务性能的关键影响。该文研究结果可为MAFB管理实践提供参考。
Abstract
[Objective] With the continuous expansion of urban rail networks and the growing complexity of passenger travel behavior, the demand for feeder bus services has become increasingly dynamic and time-sensitive. Traditional fixed-route feeder services, constrained by static schedules and rigid capacity allocation, struggle to adapt to real-time demand fluctuations during peak and off-peak periods. This mismatch between service supply and demand often results in long passenger waiting times, vehicle idleness, and operational inefficiency. To address these challenges, this study introduces modular autonomous feeder buses (MAFBs), an innovative transit service that employs modular autonomous vehicles capable of forming in real-time to flexibly meet fluctuating passenger demand. [Methods] Building on this concept, this study developed a multiobjective dynamic scheduling optimization framework within a rolling horizon approach to capture temporal variations in feeder demand. The proposed method performed periodic reoptimization within the rolling horizon, enabling MAFBs to adaptively optimize routing, scheduling, and modular capacity configurations in response to real-time demand fluctuations. At the modeling level, real-time demand information was incorporated into a unified decision-making framework, in which a multiobjective optimization formulation balanced passenger-oriented and operator-oriented objectives under dynamic demand conditions, thereby capturing the trade-offs between service quality and operational efficiency. To efficiently solve this dynamic and combinatorial problem, a multi- objective adaptive large neighborhood search (ALNS) algorithm was designed. The algorithm extended the standard ALNS by integrating multiobjective evaluation and nondominated solution maintenance mechanisms, enabling it to balance exploration and exploitation and to maintain diverse Pareto-approximate fronts within each rolling period. A case study based on Beijing’s urban rail feeder scenario validated the proposed framework. This study conducted comparative analyses between conventional feeder services and the MAFB system, followed by sensitivity experiments to investigate the effects of rolling horizon intervals and modular vehicle capacities on the overall system performance and service level. [Results] A small example demonstrated that the proposed solution approach generated well-distributed Pareto-optimal solutions.Experiments based on realistic urban feeder scenarios showed that the MAFB system achieved high-quality trade-offs between operational efficiency and service quality. Compared with conventional feeder operations, the MAFB framework reduced total operational costs by up to 23.48% while maintaining high service reliability and significantly improving passenger waiting and detour times. The rolling optimization mechanism enabled real-time adaptation to fluctuating demand, ensuring responsive system performance.Sensitivity analyses further indicated that increasing vehicle module capacity and extending scheduling intervals could further reduce costs but led to a 14.00% decrease in the average load factor and an approximately 1.40 min increase in passenger waiting time, revealing a trade-off between economy and service quality. [Conclusions] The proposed rolling horizon MAFB optimization framework effectively captures the time-varying characteristics of feeder demand and enables integrated decision-making across routing, scheduling, and dynamic capacity. By combining modular autonomous vehicle flexibility with multiobjective optimization, the framework enhances system responsiveness, operational efficiency, and passenger experience. The findings validate the feasibility of MAFBs in practical feeder operations and provide methodological insights for the development of next-generation intelligent, autonomous, and modular public transport systems, offering strong theoretical and practical support for the advancement of smart urban mobility.
关键词
接驳公交 /
模块化自动驾驶车辆 /
动态调度 /
多目标优化 /
滚动时域方法
Key words
feeder bus /
modular autonomous vehicles /
dynamic scheduling /
multi-objective optimization /
rolling horizon approach
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] LE J K, TENG J. Understanding influencing factors of travel mode choice in urban-suburban travel: A case study in Shanghai[J]. Urban Rail Transit, 2023, 9(2): 127-146.
[2] 奇格奇, 邹恺杰, 邹婕, 等. 面向异质化需求的无人驾驶电动公交接驳路径优化[J]. 清华大学学报(自然科学版), 2022, 62(7): 1178-1185. QI G Q, ZOU K J, ZOU J, et al. Feeder transit routing optimization of driverless electric buses for heterogeneous demands [J]. Journal of Tsinghua University (Science and Technology), 2022, 62(7): 1178-1185. (in Chinese)
[3] YANG Y, ZHAN J H, LIU Y, et al. Cross-city transfer learning: Applications and challenges for smart cities and sustainable transportation[J]. Communications in Transportation Research, 2025, 5: 100206.
[4] MELIS L, SÖRENSEN K. The real-time on-demand bus routing problem: The cost of dynamic requests[J]. Computers & Operations Research, 2022, 147: 105941.
[5] 任婧璇, 常孝亭, 巫威眺, 等. 考虑候选站点和全服务过程的需求响应接驳公交调度[J]. 交通运输系统工程与信息, 2023, 23(5): 202-214. REN J X, CHANG X T, WU W T, et al. Demand responsive feeder transit scheduling considering candidate stops and full-service process[J]. Journal of Transportation Systems Engineering and Information Technology, 2023, 23(5): 202-214. (in Chinese)
[6] 孙倩, 胡大伟, 钱一之, 等. 考虑车辆随机到站时间的动态需求响应型接驳公交线路优化[J]. 交通运输系统工程与信息, 2022, 22(5): 196-204. SUN Q, HU D W, CHIEN S, et al. Dynamic bus routing optimization for demand-responsive feeder transit considering stochastic bus arrival time[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(5): 196-204. (in Chinese)
[7] GALARZA MONTENEGRO B D, SÖRENSEN K, VANSTEENWEGEN P. The real-time dynamic online feeder service with a maximum headway at mandatory stops[J]. Transportmetrica A: Transport Science, 2025, 21(1): 2227738.
[8] HE D D, GUAN W, LIU T, et al. Coordinated operation of fixed- route and demand-responsive feeder transit services in a travel corridor[J]. Journal of Transportation Engineering, Part A: Systems, 2023, 149(2): 04022142.
[9] LIU X H, QU X B, MA X L. Improving flex-route transit services with modular autonomous vehicles[J]. Transportation Research Part E: Logistics and Transportation Review, 2021, 149: 102331.
[10] 刘洋, 占佳豪, 李深, 等. 自动驾驶技术的未来: 单车智能和智能车路协同[J]. 汽车安全与节能学报, 2024, 15(5): 611-633. LIU Y, ZHAN J H, LI S, et al. Future of autonomous driving: Single autonomous driving and intelligent vehicle-infrastructure collaboration systems[J]. Journal of Automotive Safety and Energy, 2024, 15(5): 611-633. (in Chinese)
[11] 何逸煦, 林泓熠, 刘洋, 等. 强化学习在自动驾驶技术中的应用与挑战[J]. 同济大学学报(自然科学版), 2024, 52(4): 520- 531. HE Y X, LIN H Y, LIU Y, et al. Applications and challenges of reinforcement learning in autonomous driving technology[J]. Journal of Tongji University(Natural Science), 2024, 52(4): 520-531. (in Chinese)
[12] 郑元, 张羽, 何蜀燕, 等. 多源时变延误下考虑信息同步的智能网联车队控制策略[J]. 交通运输工程与信息学报, 2022, 20(2): 25-41. ZHENG Y, ZHANG Y, HE S Y, et al. Control Strategy for connected and automated vehicle platoon considering information synchronization under time-varying delays of multiple sources [J]. Journal of Transportation Engineering and Information, 2022, 20(2): 25-41. (in Chinese)
[13] KHAN Z S, MENÉNDEZ M. Bus splitting and bus holding: A new strategy using autonomous modular buses for preventing bus bunching[J]. Transportation Research Part A: Policy and Practice, 2023, 177: 103825.
[14] LIU Z K, DE ALMEIDA CORREIA G H, MA Z L, et al. Integrated optimization of timetable, bus formation, and vehicle scheduling in autonomous modular public transport systems[J]. Transportation Research Part C: Emerging Technologies, 2023, 155: 104306.
[15] JI Y X, LIU B, SHEN Y, et al. Scheduling strategy for transit routes with modular autonomous vehicles[J]. International Journal of Transportation Science and Technology, 2021, 10(2): 121-135.
[16] ZERMASLI D, ILIOPOULOU C, LASKARIS G, et al. Feeder bus network design with modular transit vehicles[J]. Journal of Public Transportation, 2023, 25: 100078.
[17] TANG C Y, LIU J Q, CEDER A, et al. Optimisation of a new hybrid transit service with modular autonomous vehicles[J]. Transportmetrica A: Transport Science, 2024, 20(2): 2165424.
[18] TIAN Q Y, LIN Y H, WANG D Z W, et al. Toward real-time operations of modular-vehicle transit services: From rolling horizon control to learning-based approach[J]. Transportation Research Part C: Emerging Technologies, 2025, 170: 104938.
[19] GUO R G, GUAN W, VALLATI M, et al. Modular autonomous electric vehicle scheduling for customized on-demand bus services[J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24(9): 10055-10066.
[20] AZADEH S S, ATASOY B, BEN-AKIVA M E, et al. Choice- driven dial-a-ride problem for demand responsive mobility service[J]. Transportation Research Part B: Methodological, 2022, 161: 128-149.
[21] WU W T, ZHANG Z Y, LU K, et al. Dynamic demand- responsive transit scheduling with time-dependent travel times: A joint supply and demand management approach[J]. Transportation Research Part E: Logistics and Transportation Review, 2025, 202: 104232.
[22] SUN X, GUO S S, GUO J, et al. A Pareto-based hybrid genetic simulated annealing algorithm for multi-objective hybrid production line balancing problem considering disassembly and assembly[J]. International Journal of Production Research, 2024, 62(13): 4809-4830.
[23] RIFAI A P, NGUYEN H T, DAWAL S Z. Multi-objective adaptive large neighborhood search for distributed reentrant permutation flow shop scheduling[J]. Applied Soft Computing, 2016, 40: 42-57.
[24] COTA L P, GUIMARÃES F G, RIBEIRO R G, et al. An adaptive multi-objective algorithm based on decomposition and large neighborhood search for a green machine scheduling problem[J]. Swarm and Evolutionary Computation, 2019, 51: 100601.
[25] HELLSTEN E O, SACRAMENTO D, PISINGER D. An adaptive large neighbourhood search heuristic for routing and scheduling feeder vessels in multi-terminal ports[J]. European Journal of Operational Research, 2020, 287(2): 682-698.
[26] LUST T, TEGHEM J. Two-phase Pareto local search for the biobjective traveling salesman problem[J]. Journal of Heuristics, 2010, 16(3): 475-510.
[27] LIANG X Y, CHEN H X, WANG N M, et al. Bi-objective vehicle routing for perishable products delivery with consideration of customers’ priorities and customized delivery time windows[J]. IEEE Transactions on Intelligent Transportation Systems, 2024, 25(11): 16881-16896.
基金
国家自然科学基金面上项目(72371021, 52472336)