摘要
针对共享单车不科学、不合理的资源配置与管理调度增加交通拥堵情况、阻碍自身健康有序发展的问题,采用LOF算法对北京市共享单车数据进行预处理的基础上,利用支持向量机回归预测北京不同区域的共享单车静态需求量,结合层次分析法分析居住区、教学区、商业区等不同区域共享单车静态需求数量的分配权重;考虑动态时间因素和单车转运的运输成本、建筑分布以及总体均衡等问题,建立共享单车动态需求下合理调度的双目标优化模型,并利用模拟退火算法求解,得出合理调度方案,最后给出共享单车的经济效益分析,给出相应的参考建议。
To solve the problems in the development of the shared bicycle industry such as unscientific and unreasonable resource allocation and poor management and scheduling which lead to exacerbated traffic congestion and chaotic development situation of this industry,this paper uses the LOF algorithm to preprocess the data of the shared bicycles in Beijing,predicts the static demand of shared bicycles in different regions of Beijing using support vector machine regression,and analyzes the distribution weight of the static demand of shared bicycles in the residential,teaching and commercial areas.Next,considering the factors of dynamic time and transportation cost,building distribution and overall balance in bicycle transportation,it establishes a two-objective optimization model for rational scheduling under the dynamic demand of shared bicycles which is then solved using the simulated annealing algorithm to obtain a reasonable scheduling scheme.Finally,the economic benefits of the sharing bicycles are analyzed,with the corresponding reference and suggestions given.
作者
张若菡
何颖
Zhang Ruohan;He Ying(School of Mathematics&Statistics,Hunan Normal University,Changsha 410006;School of Economics,Xuzhou University of Technology,Xuzhou 221008,China)
出处
《物流技术》
2018年第11期64-70,共7页
Logistics Technology
基金
江苏省博士后基金项目"纵向数据半变系数模型的统计推断及应用研究"(1601076B)
关键词
共享单车
资源调度
支持向量机回归
模拟退火
双目标规划
shared bicycle
resource scheduling
support vector machine regression
simulated annealing
two-objective programming