基于随机仿真优化的公交时刻表再编制  

Bus timetable reformulation based on stochastic simulation optimization

作　　者：刘鹏杰 郑亮[1] LIU Pengjie;ZHENG Liang(School of Traffic and Transportation Engineering,Central South University,Changsha 410075,China)

机构地区：[1]中南大学交通运输工程学院,湖南长沙410075

出　　处：《铁道科学与工程学报》2022年第11期3190-3198,共9页Journal of Railway Science and Engineering

基　　金：国家自然科学基金资助项目(71871227);湖南省自然科学基金资助项目(2021JJ30888);中南大学创新驱动计划资助项目(2019CX018)。

摘　　要：时刻表作为公交企业与出行乘客之间的沟通桥梁,在满足乘客需求、增强公交可靠性等方面起着至关重要的作用。然而,现有均匀发车间隔和均衡载客量的时刻表之间各有利弊且难以权衡。此外,公交系统中随机的行程时间和乘客需求使得时刻表难以充分发挥效用。鉴于此,需要建立随机规划模型,以实现在均匀发车间隔时刻表的基准上进行微调再编制从而达到减少乘客平均等待时间和均衡最大载客量的目的。针对难以求解的随机规划模型,通过Monte Carlo方法将随机规划模型转化为随机仿真优化模型。使用基于随机克里金(SKG)的多点填充采样全局优化算法进行随机仿真优化模型的求解。研究结果表明:该方法能够达到同时减少乘客平均等待时间和均衡最大载客量的目的,且在不同权重组合下总目标至少能改进8.38%,最多能改进29.68%。通过灵敏度分析研究乘客平均等待时间和均衡最大载客量之间的冲突关系。提出的基于随机仿真优化的公交时刻表再编制方法在减少乘客平均等待时间、均衡最大载客量、改善公交服务水平、提升公交吸引力方面具有丰富的研究价值与现实意义。As a communication bridge between bus companies and passengers, timetables play a vital role in meeting passenger demands and enhancing bus reliability. However, there are pros and cons between the existing even departure interval and the balanced passenger load timetables, and it is difficult to weigh them. In addition,random travel times and passenger demand make it difficult for the timetable to be fully effective. In view of this,this paper first develops a stochastic programming model to fine-tune the even departure interval timetable to reduce the average passenger waiting time and equalize the maximum passenger load. Then, for the stochastic programming model that is difficult to solve, this paper uses the Monte Carlo method to transform it into a simulation-based optimization model. Finally, the global optimization algorithm of multi-point filling sampling based on stochastic kriging(SKG) is used to solve the problem. Experiments show that this method can simultaneously reduce the average waiting time of passengers and balance the maximum passenger load, and can improve at least 8.38% and at most 29.68% under different weight combinations. Furthermore, the conflict relationship between targets is studied by sensitivity analysis. The method proposed in this paper has rich research value and practical significance in improving bus services and enhancing bus attraction.

关 键 词：公交时刻表 随机行程时间 随机乘客需求 仿真优化 

分 类 号：U491.4[交通运输工程—交通运输规划与管理]