长期以来，我国社会物流总成本占GDP比例过高的问题广受关注，而城市配送在整个物流环节中扮演的角色也越发重要，如何提高城市配送效率成为降低社会物流总成本的关键一环。本文以HL物流企业西安配送中心为研究对象，通过对该配送中心的实际调研，发现其业务流程有待改善，车辆配送系统存在明显缺陷。该配送中心对订单进行配送前，会先根据订单配送地址将所有订单分区，再对每个分区后的订单独立配送，这就导致配送结果很难达到全局最优；目前车辆配送系统的主要作用为管理基础数据，缺少车辆智能配载功能模块，所有的配车操作需要调度员根据工作经验手动完成，以至配送结果往往差强人意。这些问题使得配送中心的成本居高不下，甚至出现在订单旺季人手不足，订单淡季人手冗余的现象。

本文首先研究了车辆路径问题及其求解算法，其次通过分析HL物流企业西安配送中心的业务现状，建立以配送车辆行驶路程最短为目标的车辆路径优化模型，并使用遗传算法求解该模型，同时使用Java语言实现遗传算法，开放接口供车辆智能配送系统调用。通过分析该配送中心整体系统的现状，结合调用遗传算法接口所需的参数，对车辆智能配送系统进行功能性和非功能性需求分析，并根据系统设计目标和原则进行系统功能设计和数据库设计。在开发车辆智能配送系统的过程中，选用JSP作为前端展示页面，JavaScript作为前端脚本语言，同时采用了jQuery和EasyUI两大主流前端框架协助开发；系统后端选用基于Java语言的SSM架构和MySQL数据库进行设计开发。

最终，成功开发出以智能配载为核心功能的车辆智能配送系统，在对配送中心基础数据管理的基础上实现了快速高效地自动生成最优的车辆配送路径，并通过HL物流企业西安配送中心某日的实际订单数据，验证了系统能够很好地提高配送效率，降低配送成本，为企业带来一定的经济价值。

For a long time, the problem of the high proportion of total social logistics cost to GDP in China has attracted wide attention, and the role of urban distribution in the entire logistics link has become increasingly important. How to improve the efficiency of urban distribution has become a key part of reducing the total cost of social logistics. This article takes the HL logistics company Xi'an Distribution Center as the research object. Through the actual investigation of the distribution center, it is found that its business processes need to be improved and the vehicle scheduling system has obvious defects. Before the distribution center distributes the orders, it will divide all orders first according to the order delivery address, and then distribute the orders for each partition independently, which makes it difficult to achieve a global optimal. At present, the main role of the vehicle distribution system is to manage basic data, and the lack of intelligent vehicle dispatching function leads to all vehicle distribution operations that need to be completed manually by the dispatcher based on work experience, so the delivery results are often unsatisfactory. These problems make the cost of the distribution center too high, and even the shortage of staff during the peak order period, and the phenomenon of redundant staff during the low order period.

In this paper, firstly, studies the vehicle routing problem and its solution algorithm. Secondly, it analyzes the business status of the HL logistics company's Xi'an distribution center to establish a vehicle routing optimization model with the shortest travel distance of the distribution vehicle, and uses genetic algorithm to solve the model. While using Java language to implement genetic algorithm, open interface for vehicle intelligent distribution system to call. By analyzing the current status of the overall system of the distribution center, combined with the parameters required to call the genetic algorithm interface, the functional and non-functional requirements of the vehicle intelligent distribution system are analyzed, and then system functions and databases are designed according to system design goals and principles. In the process of developing the vehicle intelligent distribution system, JSP was used as the front-end display page, JavaScript was used as the front-end scripting language, and two popular front-end frameworks, jQuery and EasyUI, were used to assist development. The back-end of system is designed and developed based on Java language-based SSM architecture and MySQL database.