【作者】 张硕；

【导师】 孟凡超；

【作者基本信息】 哈尔滨工业大学 ， 计算机技术（专业学位）， 2019， 硕士

【摘要】 随着手机购物、电子商务的发展,网上购物的数量愈加增多,这给物流终端的快件配送带来了更大的挑战。针对目前终端配送存在信息利用率低、可控性差、灵活性差等问题,很多物流公司已经开始变革,从传统的配送模式向众包模式进行转变。众包模式下的终端配送是利用先进的网络技术将闲置的人力资源整合以满足终端配送需求,为用户提供更好的服务。因此,在城市众包配送服务中,众包任务和配送资源的合理调度是影响配送效率的关键因素。为此,本文进行了城市众包配送服务调度方法的研究与系统实现,为众包任务和配送资源提供更好的调度方法,提高配送效率,促进众包物流行业的快速发展,具体包括以下几方面的工作:首先,构建基于众包模式的快件配送服务模型,包含众包模式的快件配送服务模式、快件配送业务模型和资源管理模型,分析众包配送中的参与者的主要功能需求及非功能需求,明确系统中角色之间的关系和主要功能,并阐述关键业务流程。其次,针对现今城市众包资源匹配研究不精准的问题,提出城市众包资源匹配模型及求解方法,定义城市众包资源匹配概念,分析众包任务的属性特征和配送资源的属性特征,然后构建城市众包资源匹配模型,并对“点我达公司”提供的真实数据进行处理,利用神经网络、SVM、朴素贝叶斯等机器学习方法进行求解,最后对资源匹配实验结果进行对比分析。再次,针对城市众包配送服务调度问题,提出基于图搜索的众包配送服务调度方法,将基于路网的任务分配问题映射成基于图结构的任务分配问题,构建最多数量分配的任务分配模型,根据路径规划提出基于任务位置的众包任务分配算法和基于蚁群规划的众包任务分配算法,生成仿真实验数据,设计实验并对比分析实验结果。最后,在上述工作的基础上,设计城市众包配送服务调度系统的总体架构、功能模块和数据库,实现城市众包配送服务调度系统,并完成调度算法的功能验证和测试。更多还原

【Abstract】 With the development of mobile shopping and e-commerce,the number of online shopping is increasing,which brings greater challenges to the delivery of logistics terminals.In view of low utilization rate,poor controllability and poor flexibility problems in terminal distribution,many logistics companies have begun to change from traditional distribution mode to crowdsourcing mode.Crowdsourcing mode of terminal distribution uses advanced network technology to integrate idle human resources and meet terminal distribution requirements to provide better services for users.Therefore,in urban crowdsourcing distribution services,the rational scheduling of crowdsourcing tasks and distribution resources is a key factor affecting the distribution efficiency.For this reason,this paper studies and implements the urban crowdsourcing distribution service scheduling methods,provides better scheduling methods for crowdsourcing tasks and distribution resources,improves distribution efficiency,and promotes the rapid development of the crowdsourcing logistics industry,including the following aspects:Firstly,the express delivery service model based on crowdsourcing mode is built,including crowdsourcing mode of express delivery service mode,express delivery business model and resource management model.Analyze the main functional and non-functional requirements of the participants in crowdsourcing,clarify the relationship between roles and main functions in the system,and explain the key business processes.Secondly,aiming at the inaccurate research of urban crowdsourcing resource matching,this paper proposes the urban crowdsourcing resource matching model and solving methods,define the concept of urban crowdsourcing resource matching,analyze the attribute characteristics of crowdsourcing tasks and distribution resources,then construct the urban crowdsourcing resource matching model.The real data provided by “Dianwoda Company” is processed,and the feature classification algorithms such as Neural Network,SVM and Naive Bayes are used to solve the problem.Finally,the results of resource matching experiment are compared and analyzed.Thirdly,aiming at the problem of urban crowdsourcing distribution service scheduling,this paper proposes a crowdsourcing distribution service scheduling method based on graph search.The task assignment problem based on road network is mapped to the task assignment problem based on graph structure.The task assignment model of the maximum quantity assignment is constructed.A crowdsourcing task assignment algorithm based on task location and a crowdsourcing task assignment algorithm based on ant colony programming are proposed according to path planning.The simulation experimental data is generated,the experiments are designed and the algorithm results are compared and analyzed.Finally,based on the above analysis and design,the architecture,function modules and database of the urban crowdsourcing distribution service dispatching system are designed,the urban crowdsourcing distribution service dispatching system is implemented,and the function verification and scheduling algorithm test are completed.更多还原