共享单车调度模型及算法研究  

Scheduling Model and Algorithm for Shared Bicycle

作　　者：于德新[1,2,3] 张行 王薇 邢雪[1] 刘珩 YU Dexin;ZHANG Hang;WANG Wei;XING Xue;LIU Heng(School of Transportation,Jilin University,Changchun 130022,Jilin,China;Jilin Intelligent Transportation Engineering Research Center,Changchun 130022,Jilin,China;Jilin Province Key Laboratory of Road Traffic,Changchun 130022,Jilin,China)

机构地区：[1]吉林大学交通学院,吉林长春130022 [2]吉林省智能交通工程研究中心,吉林长春130022 [3]吉林省道路交通重点实验室,吉林长春130022

出　　处：《重庆交通大学学报（自然科学版）》2020年第7期1-7,14,共8页Journal of Chongqing Jiaotong University(Natural Science)

基　　金：国家自然科学基金资助项目(51308249);山东省省管企业科技创新项目(20122150251-1)。

摘　　要：共享单车运营过程中出现的周转率低、调度成本过高等问题制约了城市慢行交通的发展。针对现有自行车调度模型假设条件的局限性,提出共享经济模式下转运最大化策略。在详细分析共享单车调度成本及相关参数的基础上,以成本最小和投放率最高为目标建立了共享单车调度模型。引入精英策略和进化算子对遗传算法进行改进,并采用TOPSIS法在改进算法求解出的有效路线集中选择最优路线。选取北京市某区域进行仿真分析,结果表明改进算法寻优能力较好。与常规调度模型相比,可选择调度路线增加了74.3%,平均调度成本进一步降低了18.3%.The development of urban slow traffic has been restricted by the problems such as the low turnover rate and the high scheduling cost in the process of shared bicycle operation.Aiming at the limitation of the assumptions of the current bicycle scheduling model,a transfer maximization strategy in the mode of shared economy was proposed.Based on the detailed analysis of the shared bicycle scheduling cost and related parameters,a shared bicycle scheduling model was established with the goal of minimizing the cost and maximizing the launch rate.The elite strategy and evolution operator were introduced to improve the genetic algorithm.The TOPSIS method was used to select the optimal route from the effective routes solved by the improved algorithm.A certain area in Beijing was selected for simulation analysis.The results show that the improved algorithm has good ability of searching for optimal solutions.Compared with the conventional scheduling model,the alternative scheduling routes are increased by 74.3%,and the average scheduling cost is further reduced by 18.3%.

关 键 词：交通工程 城市交通 调度方法 共享单车 车辆路线问题 遗传算法 

分 类 号：U491[交通运输工程—交通运输规划与管理]