【作者】 李国旗；

【导师】 张锦；

【作者基本信息】 西南交通大学 ， 物流工程， 2010， 博士

【摘要】 物流设施的布局优化分析是物流规划工作的重要组成部分,涉及物流设施的需求分析、层次划分、数量确定、规模确定、选址、物流需求分配等一系列问题,物流设施系统规划设计的科学性、有效性决定着物流网络的运行效率和服务水平。城市物流设施是由满足城市社会经济发展、人民群众生活所需而提供物流相关功能和组织物流服务的场所,包括物流园区、物流中心、配送中心以及各类运输枢纽、货运场站、仓库等。城市物流设施布局优化的主要任务是依据城市物流需求总量以及物流需求的分方向、分运输方式、分品类分布特征,运用科学方法,确定城市物流设施的层次与类别,不同层次的物流设施的数量、规模、位置以及服务对象。由于城市物流设施的布局优化受到物流需求、城市用地条件、城市区位条件等多种因素的影响,不同类型的城市物流设施具有不同的服务半径、服务能力、服务对象等多种属性,不同类型的城市物流设施之间还具有相似、关联等综合属性,因此,研究多属性特征下的城市物流设施布局中的优化问题,对提高城市物流规划的科学性、可操作性具有积极意义,值得深入研究。本文在现有研究基础上,从城市物流设施的具有的核心属性、外延属性、综合属性入手,重点研究了城市物流设施布局优化过程中存在的分类与分层、数量与规模确定、位置确定问题。本文所做的主要工作如下：1.在归纳总结已有城市物流设施各种技术属性、特性的基础上,提出城市物流设施的属性由核心属性、外延属性、综合属性构成,并对各个具体属性的定义、计算公式进行了详细分析,着重研究了城市物流设施间存在的关联度、相似度属性。运用复杂网络理论,对城市物流设施的外延属性,即城市物流网络进行了复杂特性分析,运用聚类分析方法,对城市物流设施的核心属性,即物流设施的类别与分层提出了确定方法。2.研究了城市物流设施的数量与规模确定问题。首先,在分析已有城市物流设施的数量与规模解决方法的特性基础上,找出了已有方法的薄弱环节,提出利用城市物流设施间存在的相似度属性,构建了城市物流设施数量与规模确定的类比模型。其次,提出利用相似度法、函数拟合法、面积差法三种方式来进行相似度的确定。第三,对模型中相似元参数提出综合运用灰关联理论、粗集约简确定,指标权重参数运用基于属性依赖度的权重处理方法确定。3.研究了城市多层次物流设施的选址问题。首先考虑城市物流设施选址具有多层次、关联性、空间局限性等特征,构建了具有不同层次的物流设施服务半径约束的多层次物流设施选址的混合整数非线性规划方法模型,运用遗传算法进行求解,设计了符合解的性质的编码结构,进行了数值仿真分析。其次,针对上述模型中服务半径确定、不同层次的物流设施备选点的确定问题,提出一种运用物流需求区划分、物流设施相似度判定以及聚类分析的启发式求解算法,通过实例分析,验证了算法了便捷性、实用性。第三,考虑物流设施具有服务能力约束、物流设施间具有关联度属性,将上述模型扩展为具能力约束和关联约束的多层次物流设施选址模型,提出运用遗传算法和粒子群算法进行求解,并进行了数值仿真分析,验证了算法的有效性,并与具有不同层次物流设施服务半径约束的多层次物流设施选址模型的求解结果进行了比较。4.研究了城市应急物流设施的选址问题。首先,考虑最大化应急物流需求点的满意度,构建了无供给能力限制的应急物流设施选址模型和有供给能力限制的应急物流选址模型。其次,对上述两类模型进行拓展,应急选址从已有物流设施中选优向新建应急物流设施转变,并将难以定量的应急物流选址影响因素,利用多属性决策理论进行综合排序,形成决策偏好,然后基于多目标规划方法构建综合考虑应急物流需求点的满意度和设施建设成本,具有服务数量约束的应急物流设施选址模型,运用模拟退火算法进行求解,并进行了数值分析。更多还原

【Abstract】 The layout optimization of logistics infrastructures is an important work of the process of logistics planning.The main contents include demand analysis, hierarchical division, quantity determination, scale determination, location, distribution of logistics demand, etc. The science and effectiveness in the process of the logistics infrastructures system planning and design determine the operation efficiency and service level of the logistics network.As a place to provide logistics functions and organizational logistics services, the urban logistics infrastructures aim to meet urban social economic development and people’s living needs, however, their representation are logistics park, logistics center, distribution center, all kinds of transportation hub, freight station, warehouse, etc. This dissertation analyzes the total logistics demand, the distribution characteristics of directions of flow, modes of transportation, categories of goods. Various valid measure and methods are adopted to deterimine the category and level of the urban logistics infrastructures and quantity, scale, location service objects of different levels of the urban logistics infrastructures.The layout optimization of the urban logistics infrastructures are affected by logistics demand, the conditions of urban land and urban geographic location and other factors. Different categories of the urban logistics infrastructures have different service radiuses, service capabilities and service objects.The relations can be measured by comprehensive attributes of similarity, relevance. All these relatons are boiling down to multiple attribute problem.It is has a positive meaning in improving the scientificity and operability of the urban logistics planning.Based on existing research results, this dissertation first presents the core, extension and comprehensive attributes of the urban logistics infrastructures. The study focuses on stratification and classification, quantity and scale, location in the process of layout optimization of the urban logistics infrastructures.Overall, the main contents of this dissertation are as follows:1.Based on the existing various technical properties and characteristics,the attributes of the urban logistics infrastructures consist of the core, extension attributes and comprehensive attributes. The definition calculation formulas of each specific attributes are discussed respectively. The comprehensive attributes of similarity, relevance are emphatically discussed. Considering the extension attributes, the complex characteristics of the urban logistics network are analyzed by the complex network theory. Considering the core attributes, the determination method of the category and level are suggested by the clustering method.2.The determination methods of quantity and scale are studied. First, considering the charactersitcis of existing research methods, weak links of existing research methods are found. Then, an analogy model is proposed based on the similarity attribute. The methods of similarity degree, function fitting and area difference are proposed to determine the similarity. Finally, the parameter of the similarity unit in the analogy model is obtained by the grey relation theory and the rough sets reduction. The parameter of the index weight is obtained by the attribute dependability.3.The location problems of the multi-level urban logistics infrastructures are studied. First, considering the charactersitcis of mutiple level, relevance, space limitations in the process of location, the multi-level urban logistics infrastructures location model with constraints of service radius (M-ULIL-SR) is developed based on mixed integer nonlinear programming (MINLP) approach. The coding structure in accordance with the property of solution are designed, the model (M-ULIL-SR) is solved by the genetic algorithm (GA). A numeric example was presented to demonstrate the validity of the model.Secondly, in order to determine the service radius and the alternative points of different levels of logistics infrastructures, a heuristic algorithm with logistics demand area division, similarity judgement and clustering analysis is designed. Through a case study, the convenience and practicality of the algorithm can be verified. Finally, considering the constraints of service capacity, the attribute of relevance between logistics infrastructures, anothor model (M-ULIL-CR) is developed based on the extention of M-ULIL-SR. The solving strategy is proposed, including genetic algorithm (GA) and Partical Swarm Optimization (PSO). the correctness and effectiveness of this algorithm are proved through experiment and analysis.the results of M-ULIL-CR are compared with M-ULIL-SR.4.The location problems of the urban emergency logistics infrastructures are studied. First, considering maximize the satisfaction of emergency logistics demand points, the urban emergency logistics infrastructures location models with capacity constraints (UELIL-CS) and uncapacitated constraints (UELIL-UCS) are developed based on mixed integer nonlinear programming (MINLP) approach. Secondly, the location of the object is shiftted from the existing logistics infrastructures to the reconstructing emergency logistics infrastructures. The factors affecting the emergency logistics location with difficult to quantify are rank the alternatives by the multi-attribute decision-making theory and form decision-making preferences, the urban emergency logistics infrastructures location models with the constraints of service quantity(UELIL-SQ) is developed based on multi-objective programming (MOP) approach. The objective of UELIL-SQ are both for maximizing the satisfaction of emergency logistics demand points and minimizing the construction cost of emergency logistics infrastructures. The solving strategy is proposed by simulated annealing (SA), the effects of the model are estimated by using numerical simulation method.更多还原