基于双层规划的公共电动自行车租赁点选址模型  

Location Model for Public Electric Bicycle Rent Based on Bi-Level Programming

作　　者：胡郁葱[1,2] 陈枝伟 黄靖翔[1,2] 

机构地区：[1]华南理工大学土木与交通学院,广东广州510640 [2]现代城市交通技术江苏高校协同创新中心,江苏南京210000

出　　处：《华南理工大学学报（自然科学版）》2017年第4期118-123,共6页Journal of South China University of Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(51408237);国家级大学生创新创业训练计划项目(201510561102)~~

摘　　要：建设公共电动自行车系统是有效利用资源、满足居民经济低碳出行需求的可行途径之一.文中提出了公共电动自行车网络系统的概念,并采用出行链对其进行分析.在此基础上,建立双层规划模型解决该网络系统中的租赁点选址问题.上层模型考虑政府的目标,为系统最优模型;下层模型考虑用户的目标,为用户均衡模型.上层模型采用混合粒子群优化算法求解,下层模型采用Frank-Wolfe算法求解.算例结果表明,文中算法能够同时确定站点选址和站点规模,收敛性能较好.Developing public electric bicycle systems is a feasible way to making full use of resources and meeting people’s increasing demand for low-carbon trips. This paper first proposes a public electric bicycle network system based on trip chain analysis, and then establishes a bi-level programming model to solve the station location prob-lem of such systems. In this model, the upper part considers the government^ goal for system optimization, while the lower part shows respect for system users^nterests and user equilibrium. The upper model is solved by using the hybrid particle swarm optimization algorithm and the lower one is solved by using Frank-Wolfe algorithm. Numeri-cal results show that the proposed programming model can determine both the optimal location and the size of sta-tions in the public electric bicycle network system with good convergence.

关 键 词：公共电动自行车 选址模型 双层规划 混合粒子群优化 

分 类 号：U491.2[交通运输工程—交通运输规划与管理]