节点文献

林副产品配送优化辅助决策模型及GIS集成研究

Study on Model and GIS Integration of Assistant Decision for Forest By-products Distribution Optimization

【作者】 郭建宏

【导师】 彭道黎; 欧阳钟辉;

【作者基本信息】 北京林业大学 , 森林经理学, 2008, 博士

【摘要】 物流配送是现代物流一个非常重要的内容,它是现代市场经济体制、现代科学技术和现代物流思想的综合产物。在配送过程中,配送线路合理与否对配送速度、成本、效益影响很大。合理的配送可以有效降低物流成本,满足消费者定点、定时、定量的商品消费需求,同时减小配送对城市环境造成的交通堵塞、噪音、污染等负外部效应。目前,在物流配送业务较多的情况下,单纯采用传统的人工办法在有限的市场许可范围内处理每日的配送最优方案已是不可能的。本文从基于GIS的空间信息管理与应用为出发点,在对林副产品配送辅助决策模型研究的基础上,旨在研究基于GIS的林副产品配送辅助决策系统的集成理论和方法,主要包括如下内容。(1)分析了物流配送的发展及研究现状以及基于GIS的物流配送系统的国内外应用及研究现状,指出随着GIS/GPS技术的发展和应用,对车辆路线问题展开面向应用的研究将具有现实意义。但目前还缺少物流配送优化模型与GIS系统的集成方面的研究,同时还缺乏GIS实时路径获取及实时配送决策方面的研究和应用。(2)系统研究了物流配送优化理论及算法基础。其中物流道路网络数据的计算机表达是实现物流配送优化的基础,分析了三种主要存储结构的空间和时间复杂度以及搜索机制。最短路径问题是GIS实时路径计算的基础,研究分析了当前各种最短路径算法,并对各种算法的实现过程进行了设计,阐述了物流交通网络中两类最短路径问题的快速实现技术。车辆优化调度问题是物流配送优化的重要理论基础,研究分析了三类配送优化问题;主要包括定位——配给问题(LAP)、车辆路线安排问题(VRP)和定位——运输路线安排问题(LRP)的主要特征及实现策略。分析了车辆优化调度问题的构成要素及类型,界定了本文研究的主要问题为单物流中心非满载车辆且封闭式的配送车辆路线优化问题。启发式算法是解决车辆优化调度问题的主要方法,分析了启发式算法理论基础、策略以及求解过程。这些分析保证了本研究问题的系统完整性,有的可以指导实践,有的为模型的建立和算法提供了基础和前提。(3)分析了林副产品配送的特点和模式,指出非满载车辆优化调度问题是最具有普遍意义的研究问题。利用区位优势原理,构建了林副产品分区配送优化模型,并用遗传算法进行了模型的仿真计算,结果表明模型具有较好的适应性,能够解决林副产品配送车辆的优化调度问题。(4)研究了物流配送优化模型与基于GIS的空间信息的系统集成问题,指出在关系数据库(RDBMS)中统一存储和管理空间数据和属性数据,是比较理想的空间数据和属性数据集成方法。分析了基于SuperMap GIS的电子地图数据组织方案,并在SuperMap GIS平台上进行了泉州市电子地图的组织开发。在SuperMap GIS道路网络数据拓扑处理的基础上,设计了适应快速搜索的物流配送优化系统数据组织方案。即用两个数组来存储及管理交通网络数据,一个用来存储和弧段相关的数据(Arc_List),另一个则存储和顶点相关的数据(Node_Index)。由于Arc_List记录了弧段的双向信息,虽然开支了一定的存储空间,但提高了搜索效率。设计了物流节点(即需求点或配送点)的数据的组织方式,用一个节点数组(L_Node)和一个弧段成本矩阵(L_Arc)来单独组织物流节点数据。最后分析了基于GIS的物流配送系统功能集成的主要模式,指出从开发成本和跨平台等方面考虑,J2EE、EJB和XML组合的功能和作用将是基于GIS的物流配送系统走互联网架构的理想选择。(5)进行了基于GIS的林副产品收配一体化辅助决策系统方案的设计,提出了以订单管理为中心的林副产品收配一体化辅助决策系统通用解决方案。

【Abstract】 The distribution is a very important part of the modern logistics system. It is also an integration of the modern marking economic regimes, the modern scientific technologies and modern logistics management ideas. The reasonability of distribution route greatly impacts the speed, the cost and the efficiency of the distribution process. The distribution cost could be cut down, and the consumer demand of fixed place, time and quantity could be satisfied if the distribution process was reasonable. And the negative external effect of the urban environments such as traffic jam, noise and pollution could also be lowered. This paper based on the management and application about spatial information of GIS and study on model of assistant decision for forest by-products optimization distribution. The main intentions were to study the integration theory and methods of assistant decision system for forest by-products based on GIS. This paper mainly included the follow content.(1) The present development and study status about the logistics distribution and application and study status about its management system based on GIS was analyzed firstly. The result showed that study on being oriented to application about vehicle route optimization problem was becoming more important with developing and applying on GIS and GPS. Now the comprehensive study on integration between the model of logistics distribution optimization and GIS system was absent. And there was also lack of the study and application on real time obtaining route and making a strategic decision on distribution.(2) Basis of theory and algorithm was analyzed in a systematic way. The computer memorizing of logistics distribution network is very important to achieve the optimization of logistics distribution. So the complexity and the search mechanism of spatial and time of the three major storage structure were analyzed. The shortest path problem is the basis for GIS to calculate the real-time path. Some current shortest path algorithms were studied and the algorithmic processes of them were designed. The rapid realization methods of two types of shortest path problem of logistics transport network were described. They were the shortest path problems between two points and muliti-points. The vehicle scheduling problem is the important theoretical foundation for the logistics and distribution optimization. The main characteristics and implementation strategy of three types of distribution optimization problem were analyzed. They mainly included the Location Allocation Problems (LAP), Vehicle Routing Problems (VRP) and Location Routing Problems (LRP). The elements and types of vehicle scheduling problem were analyzed firstly. Then the main problem of this paper to study was decided as the urban logistics distribution vehicle route optimization problem based on one center, not-fully load and non open. The heuristic optimization strategy is the main methods to solve the vehicle scheduling problem. So its rationale, implementation tactics and process was analyzed. These analyses ensured integrality of study objective. Some of them might applied directly and some of them were the basis for the following models and algorithms(3) The characteristics and modes of forest by-products logistics distribution were discussed and the results were pointing out that the non-load vehicle scheduling problem is the most universal significance to the logistics distribution problem. Then according to the principle of regional advantages, an optimization model for forest by-products distribution was established. The simulation computed by genetic algorithm showed that the model has better adaptability and could solve forest by-products logistics distribution vehicles optimal scheduling problems.(4) The problems of system integration between logistics distribution optimization model and the GIS-based spatial information system were discussed. The results showed that the effective data integration approach was storing and managing the spatial data and attribute data together in the relation database management systems (RDBMS). The method of data organization about SuperMap GIS was analyzed and the electronic map of Quanzhou city was designed based on SuperMap GIS platform. Then a new data organization method of logistics distribution optimization system based on Fast Search strategy was put forward. That was using two arrays to store and manage traffic network data, one of them (Arc_List) was to store the arcs data, and the other (Nodelndex) was to store related node data. Although some extra storage spaces were used, the search efficiency was enhanced because the two-way information of arcs was recorded with Arc_List. The data organization method of the logistics nodes (also called logistics demand or distribution points) was also designed. The data was organized separately by a node array (L_Node) and an arcs cost matrix (L_Arc). The main models to integrate the functions of GIS-based logistics distribution system were discussed. They mainly included symmetric structure integration model, OLE technology integration model, COM technology integration model, XML technology integration model, CORBA technology integration model, agent technology integration model and JAVA technology integration model. The result showed that function combined by J2EE, EJB and XML was positive selection to develop the Web structure system of logistics distribution based on GIS if the cost and cross-platform was considered.(5) The scheme about support system of GIS-based forest by-products for gathering and distributing was designed and a general scheme of forest by-products distribution based on order management was proposed.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络