基于车联网及云计算的电动物流车智能调度算法  

An Electric Logistics Vehicle Intelligent Scheduling Algorithm Based on Internet of Vehicles and Cloud Computing

作　　者：赵建峰[1] 袁细国[2] 梁伯栋[1] 陈球霞 ZHAO Jian-feng;YUAN Xi-guo;LIANG Bo-dong;CHEN Qiu-xia(School of Automobile & Transportation Engineering, Shenzhen Polytechnic, Shenzhen Guangdong 518000, China;School of Computer, Xidian University,Xi’an Shaanxi 710071,China)

机构地区：[1]深圳职业技术学院汽车与交通学院,广东深圳518000 [2]西安电子科技大学计算机学院,陕西西安710071

出　　处：《公路交通科技》2019年第6期112-124,共13页Journal of Highway and Transportation Research and Development

基　　金：2018深圳市孔雀技术创新项目(KQJSCX20180329191021388);2016深圳市科创委知识创新计划(JCYJ20160429145314252,JCYJ20160527162817715,JCYJ20160407160609492);2016年广东省省级科技计划项目(2016A010101039);2015年深圳职业技术学院重点科研项目(601522k30007);深圳职业技术学院校青年基金(601522K30015)

摘　　要：考虑到轻型电动货车作为未来城市内物流运输的主要载体,以及云计算和车联网在物流行业的应用,在对物流企业调研的基础上,研究了未来电动车作为城市货运物流的调度问题。区别于已有研究成果将车辆装配与路径规划分开进行优化的研究思路,基于未来物流企业将普及云计算平台及车联网技术的假设,构建了包含货物装配及车辆路径规划一体的调度模型。根据企业物流调度的实际需求,改变了以往以单一节点为中心的路网结构,构建了更加符合实际的全连通路网结构。提出采用平均道路运输成本、平均车辆装卸成本、仓库的仓储成本、仓储的均衡度,货物运输的剩余时间等5个量化评价指标对调度结果的优劣进行评价;在调度建模的基础上,提出了一种新型实用的基于车联网及云计算平台的电动车物流的多目标优化调度算法,用于对调度模型的求解。为验证模型的有效性及算法正确性,生成了不同规模的数据集进行测试。首先在小规模数据上验证了模型与算法的正确性,然后在大规模不同调度请求下,对比智能调度算法与当前物流企业普遍采用的人工调度算法,在不同仓库的仓储能力与车辆的运输能力的比值、不同调度车辆数量、不同仓储节点数量下的调度情况。100组随机数据的平均调度结果分析表明:智能调度算法调度指标均优于人工调度算法。Considering the light electric truck will be the main carrierof future urban logistic transport and the application of cloudcomputing and internet of vehicle in the logistic industry, thescheduling problem of future electric vehicle for urban freightlogistics is studied based on the review of logistics enterprises.Different from existing research results of optimizing vehicle assemblyand path planning separately, a new scheduling model which includes boththe cargo assembly and vehicle path planning is proposed based on theassumption that logistics enterprises will become a popular cloudcomputing platform as well as a popular vehicle networking technology inthe future. According to the actual demand of enterprise logisticsscheduling, a more realistic fully connected network other than thetraditional single-node-centric network structure is constructed.Moreover, 5 quantitative evaluation indexes, including average roadtransport cost, average vehicle loading and unloading cost, warehousestorage cost, warehouse balance degree and remaining time of cargotransport, are proposed to evaluate the scheduling result. Based on thescheduling modeling, a new practical electrical vehicle logisticsmulti-objective optimization scheduling algorithm is proposed based oninternet of vehicle and cloud computing to solve the scheduling model.In order to verify the effectiveness of the model and the correctness ofthe algorithm, a data set with different scales is generated. First, thecorrectness of the model and the algorithm is verified by a set of smallscale data. Then, with different large-scale scheduling requests, theintelligent scheduling algorithm is compared with the manual schedulingalgorithm generally used by logistics enterprises under the ratios ofdifferent warehouse storage capacities and vehicle transport capacities,different quantities of dispatched vehicles and storage nodes. Theaverage scheduling result of 100 groups of random data shows that theintelligent scheduling algorithm is superior to the manual schedulingalgori

关 键 词：城市交通 智能调度算法 多目标优化算法 物流车辆调度 车联网 云计算 电动车 

分 类 号：U491[交通运输工程—交通运输规划与管理]