【作者】 宋杨；

【导师】 王立海；

【作者基本信息】 东北林业大学 ， 森林工程， 2010， 博士

【摘要】 随着国际金融危机对我国实体经济影响的逐步加深,物流业作为重要的服务产业也受到了严重冲击。同时,物流企业发展还很不成熟、数量众多而规模偏小；多数物流企业服务集中于专业化的物流运作,功能层次较低的现状严重制约了我国现代物流产业的发展。从供应链管理的角度,对若干物流企业进行有效的资源整合,进而提高整个物流行业的服务质量应是促进我国物流产业发展的重要对策。而第四方物流恰恰是实现物流资源整合的有效形式。由于第四方物流系统的有效运作需要供应链上的节点企业,物流企业、第四方物流企业的动态协作来实现,因此运用复杂适应系统理论对第四方物流系统的运作问题进行研究是必要而可行的,而在此基础上研究促进第四方物流系统物流资源整合过程中的合作伙伴选择、物流任务分配机制和任务分配后的动态协调等关键问题的机制和优化方法,具有重要的理论意义和现实意义。本文在对复杂适应系统理论、供应链管理理论、遗传算法等三个方面内容进行阐述的基础上,主要对五个方面的问题进行了研究。首先对第四方物流系统的概念进行了界定,提出了该系统的三类要素：供应链上的节点企业、物流企业和第四方物流企业并针对复杂适应系统理论提出的所有复杂适应系统所共有的七个基本特征逐一同第四方物流系统的实际情况进行对比,提出了第四方物流系统属于复杂适应系统的观点,并提出了系统运作过程复杂性、资源整合形式虚拟性和资源整合动态性等三个第四方物流系统的个性特征。其次,本文将第四方物流系统中要素的概念转化为复杂适应系统理论中主体的概念,提出了第四方物流系统中的三类主体：4PLAgent、LAgent和SCCAgent,并对各类主体中的基于功能的子主体进行了分析,提出了第四方物流系统动态运作模型并以复杂适应系统理论中机制的内涵(机制是用转换函数定义的规则及规则间相互作用规则的总称)为基础提出了第四方物流系统的三项关键运作机制：物流合作伙伴选择机制、物流任务分配机制和基于任务分配的动态协调机制。第三,以《物流企业分类评估标准》为基础,针对第四方物流系统的特点和物流合作伙伴选择的影响因素,提出了物流合作伙伴选择指标体系并应用层次分析法对指标体系的权重系数进行了研究。鉴于第四方物流企业往往选择多个类型的多个物流企业结成合作伙伴关系的事实,提出了基于备选物流企业综合情况比较、按区域划分物流需求的4PL模式下物流合作伙伴选择的优化模型并运用遗传算法,通过采用二进制编码、较大初始种群、较小交叉概率和较大变异概率等方法对物流合作伙伴选择问题进行了求解。第四,在分析影响4PLAgent任务分配的三方面因素,即LAgent物流服务的综合实力、物流服务质量及任务分配后的满意度的基础上,提出了4PLAgent物流任务分配的四点规则和物流任务分配问题的多目标优化模型。由于目标选择不同,物流任务分配结果不同。本文提出了以物流服务质量最优(基于SCCAgent的)和物流服务质量最优、LAgent满意度最大(基于4PLAgent的)的两种物流任务分配机制并并运用遗传算法,通过采用自然数编码的方式,采用惩罚函数、较大初始种群、较小交叉概率和较大变异概率的方法,对物流任务分配问题进行了求解。最后,构建了“基于SCCAgent"和“基于4PLAgent"的两种物流任务分配机制的层层递进的三个动态模型并运用MATLAB7.0软件对动态模型进行仿真,仿真过程中,通过调整基础数据和参数得到了具有说服力的仿真结果,说明“基于4PLAgent"的物流任务分配机制优于“基于SCCAgent"的物流任务分配机制,综合考虑SCCAgent与LAgent满意程度的物流任务分配机制应成为第四方物流系统任务分配的首选机制。更多还原

【Abstract】 With the international finance crisis influence on real economy being larger and larger in china, Logistics industry which is an important service industry is impacted severely too. At the same time, the reality of small scale and single service function of numerous logistics enterprises has greatly restricted the development of China’s modern logistics industry. The integration of the resources of logistics enterprises from the angle of supply chain management that can enhance the service quality should be the important countermeasure that will accelerate the development of logistics industry. And the fourth party logistics (4PL) is a good form of the integration of the logistics resources. Because the achievement of the aim of the 4PL system needs dynamic cooperation of node enterprises of supply chain and logistics enterprises and 4PL enterprise, it is necessary and feasible to research the question of 4PL operation based on the theory of CAS. And it’s both theoretically and practically meaningful to study the key mechanisms and optimization method of the questions of logistics partner selection, the logistics order allocation and the dynamic harmony that can improve the result of the logistics resources integration.In this paper, the questions of five aspects are studied on the basis of the expatiation of the theories of CAS, SCM and GA.Firstly, the Conception of 4PL System and the three kinds of elements are put forward, and by contrasting the fact of 4PL system with the seven characters that every CAS should have, the 4PL system is ascertained as a kind of CAS. Meanwhile, three individualities of 4PL system are brought forward.Secondly, three kinds of Agents including 4PLAgent, LAgent and SCCAgent are supposed, and the dissertation analyzes the sub-agent of each Agent, puts forward the dynamic operation model of 4PL system and the three core mechanisms:the mechanism of logistics partner selection, the mechanism of logistics order allocation and the dynamic harmony mechanism based on logistics order allocation.Thirdly, the dissertation proposes logistics partner selection index system and uses the method of AHP to determine weight coefficient of each index, on the basis of analysis of the content of "Logistics Enterprises Classification Assessment Standard" and the influence factors of logistics partner selection. Whereas the 4PL enterprise usually selects more than one logistics enterprise of more than one kind as its partners, the dissertation proposes an optimization model of the logistics partner selection under the 4PL model, based on comparing strength of each logistics enterprise and compartmentalizing the logistics demand by region. And an improved algorithm is presented based on genetic algorithm by using the methods of binary encoding, larger initial population size, smaller crossover probability and larger variation probability.Fourthly, four rules and an optimization model of multi-aims of logistics order allocation are brought through analyzing the influence factors of logistics order allocation. Different aim, different result, the dissertation puts forward two kinds of order allocation mechanisms named mechanism based on SCCAgent and mechanism based on 4PLAgent. And an optimization model for logistic order allocation and the improved algorithm are presented based on GA by using the methods of natural number coding, punished-function, larger initial population size, smaller crossover probability and larger variation probability.At last, the dissertation constitutes three dynamic models based on the above two kinds of order allocation mechanisms, and is used the software of MATLAB7.0 to simulate the dynamic process. The convincing simulation result is achieved through modulating the basic data and parameters during the dynamic process of emulation. The result shows that order allocation mechanism based on 4PLAgent is better than the mechanism based on SCCAgent and we should choose mechanism based on 4PLAgent to allocate the logistics order.更多还原