【作者】 艾耕云；

【导师】 陈俊芳；

【作者基本信息】 上海交通大学 ， 企业管理， 2008， 博士

【摘要】 上世纪末以来,供应链管理成为企业获取竞争优势,增强竞争力的重要手段;网络化成为企业快速响应多变市场的不可或缺的策略方法,网络合作关系的成员企业通过发展、深化、集成合作伙伴关系,彼此依赖,密不可分;降低交易成本、集中核心能力、获得关键技术、共享风险是供应链管理成功的关键。供应链给企业带来竞争优势的同时,也带来了潜在风险。人们在研究供应链网络设计问题、供应链协调与契约问题、供应链中的库存问题、供应链绩效评价问题、供应链中的信息流与价值等问题的同时,越来越关注供应链的风险问题。供应链风险管理虽然引起大家的广泛研究,然而采用非线性复杂理论把供应链作为整体进行研究的文献很少。本文尝试从分形理论角度,利用非线性分析工具,研究供应链整体风险结构和风险管理问题,作些初始的探索,本文的主要工作和研究结论如下:1、简单阐述了供应网络脆弱性、供应网络弹性、供应网络鲁棒性、供应网络柔性、供应网络可靠性、供应链敏捷性、供应网络结构、供应链复杂度等相关概念,分析其与供应链风险的关系。系统总结了供应链风险管理相关的国内外文献,对研究现状进行了简短总结。2、提出了具有分形特征的供应链优化了信息空间的维数,优化了信息效率。通过对供应链风险因素和参与方广泛性和复杂性的分析,指出供应链风险管理的巨大信息需求,为了对供应链风险管理的整体有效掌控,本文认为必须寻找有效的信息机制,提高信息效率,以最小的信息需求达成有效控制和管理。通过对机制设计大师Hurwicz门徒田国强的分散化信息机制田氏定理的剖析,可知获得最优信息效率、实现资源最优配置的经济环境类的信息空间及效用函数与生产函数具有分形的特征。因此,本文结合信息机制和分形理论的分析,提出对供应链风险结构进行重构,建立具有分形特征的供应链风险结构是解决供应链风险管理信息效率的最佳途径。3、研究了重构分形特征供应链的意义及其重构的理论与实践方法。利用分形理论,把复杂的企业和供应链系统压缩为一些自相似规则的分形体,这样可以减少信息的维数,利用较少数量的指标信息,达成管理和控制更大规模、更长链条的企业或供应链,降低供应链管理风险。先从竞争效率、激励机制和自组织机制角度研究了重构分形特征供应链的意义,其自相似性质引致了自组织、自协同和自优化,提高了系统的效率和能力。再从理论和实践两个方面探讨了供应链风险结构重构的方法。理论方面,主要以主成分分析法甄选出优秀个体的互不相关的关键因素,以此关键因素的标准和规范为核心标准源,重构具有自相似特征的供应链。实践方面,从组织结构、制度程序、校准强化系统、企业文化、员工发展、信息系统六个方面构造因素重构具有分形特征的供应链核心企业。按照统一的规则在核心企业内部与供应链的各类供应商之间及供应商的上游供应商之间进行渗透和扩展,包括个体组织的运作采用相似机制、每个供应商选择及管理采用相似机制、外部合作伙伴多级延伸管理采用相似机制、内部组织不同层级管理采用相似机制。把供应链核心企业及其供应链系统压缩为一些相似规则的分形体。4、研究了供应链风险结构及其特征参数。分析了基于价值活动的供应链嵌套结构,提炼了结合该嵌套结构的供应链风险因素,并进一步架构了供应链风险分析指标体系。对供应链风险结构研究的理论选择进行了分析,确定分形理论能够更好地刻画供应链的风险结构特征,基于分形理论阐述了价值活动风险结构的分形维数、多级供应链风险结构的局部分形维数的计算方法。研究了供应链风险结构分形奇异谱及其特征参数,分形奇异谱刻画了供应链不同局部区域的供应链风险离差度,不规则度、不均匀程度或复杂度。进一步讨论了改变分形元尺度能够使我们从不同的尺度范围来分析和管理供应链风险结构。运用算例具体描绘了供应链分形维数、多重分形及其特征参数的计算过程,得到了特征参数的具体数值。本文使供应链风险管理由两层延伸至更多层,拓展了供应链风险管理的视野,更接近客观现实供应链风险的实际状况,建立了供应链整体的风险管理观念。5、研究了供应链风险扩散收敛模型及管理机制。基于供应链系统的复杂性和供应链核心企业内部及供应链链级之间作用的非线性,运用复杂理论的非线性模型对供应链风险扩散收敛机理进行了分析。研究了供应链成员不实施干预和实施干预、干预存在时滞而参数不存在扰动和干预不存在时滞而参数存在随机扰动四种情形下的供应链风险扩散收敛模型。模型着重分析了供应链风险因素如何在供应链网络系统中存在,其产生、传导、扩散、震荡、收敛并稳态运行的方式,供应链风险在供应链系统中扩散收敛的关键参数,即供应链风险扩散收敛系数、平均响应时间及风险管控能力系数。针对模型的三个关键参数,探讨了供应链风险结构对风险扩散收敛系数的影响,得出了供应链风险结构分形维数奇异谱的特征参数可以有效表征风险扩散收敛系数,并提出了减低供应链复杂度而降低供应链风险的管理方法;探讨了供应链敏捷性对供应链平均响应时间的影响,分析了供应链敏捷性评估模型和改善敏捷性水平的方法;探讨了不同绩效评估机制对供应链成员管控风险努力水平的影响,从而改善供应链管理者的努力程度,改进供应链风险管控能力。从分形特征供应链自相似性提高信息效用方面、自组织、自优化能力适应环境变化方面得出分形特征供应链能够兼备较高的供应链风险扩散收敛系数、较小的供应链风险响应时间,以较小的供应链风险结构变化化解风险;供应链的自相似性也有利于绩效的比较、有利于更好运用绩效评估机制;因此,分形特征供应链能够获得更优的三个模型参数水平,提高供应链的抗风险能力。6、研究了分形供应链自相似性对收益的影响。探讨了基于分形理论而重构的供应链系统,其规则和结构的相似性导致供应链组成部分之间风险具有较强的相关性。而供应链自相似性对风险具有双面的作用,自相似引致自相关产生加强风险作用,自相似引致自优化又产生降低风险的作用,两种方式彼此作用。本文借鉴Babich(2003)供应商信用风险相关性模型和Martin A. Lariviere & Evan L. Porteus(2001)纯定价合约报童问题模型,研究供应链自相似性引致的自相关和自优化对收益的综合影响,分别从供应链纵向自相似和横向自相似两个方面给予揭示,即生产商和单个供应商的纵向自相似性、生产商和多个供应商的横向自相似性,在确定性需求和不确定性需求下,对供应链各方均衡及收益的影响。单个供应商、需求不确定下,一体化供应链均衡最佳订货量和最佳利润随着产品缺陷风险减小而增大;非一体化供应链均衡最佳订货量、最佳供应商利润、最佳生产商利润均随产品缺陷概率减小而增大。因为供应链自相似导致自优化而降低产品缺陷概率,从而供应链自相似性提高有利于一体化和非一体化供应链收益的提高。两个供应商、需求确定下,当自相似引致的自优化达到一定程度时,供应链自相似性提高,使得生产商和供应链整体利润得到提高;供应链自相似性对供应商的影响不确定,只有供应商缺陷概率很高时,才可能使供应商利润提高。两个供应商、需求不确定下,通过算例分析,在需求为指数分布和正态分布,对称性供应商情形时,得出供应链自相似性导致自优化达到一定水平后,供应链自相似性的提高,使生产商和供应链整体均衡利润上升,使供应商均衡利润下降。因为供应链自相似性提高使得供应链整体利润得到提升,供应商的积极性可以通过供应链协调来提高。7、对具有分形特征的供应链风险结构及管理机制进行了实证分析。总结了自相似程度较高的连锁经营企业的发展现状,客观上证实具有分形特征的企业及供应链能够较好地适应不断变化、竞争激烈的环境,有效防范和控制风险,获得良好发展。分析了国内两家有代表性的中西餐饮企业,详细阐述其结构特点和管理原则,计算了其供应链价值活动分形维数,剖析了奇异指数和奇异谱,刻画出其供应链风险结构差异,得出了供应链扩散收敛系数的相对值。通过两家企业2002年—2007年几个重大危机事件中企业供应链实际营运情况的深入分析,得出了供应链风险管控能力系数越大,供应链越能在较低风险的稳态下运行;供应链风险扩散收敛系数的相对值虽然较大,只要相对平均响应时间短,响应速度快,供应链风险危机就能够被有效控制,恢复至稳态运行。通过多年企业实际运营数据的比较,可以发现自相似程度高的企业,虽然其导致较高的自相关性,但由于自组织、自优化能力较强,防范风险、化解风险能力也较强,自相似性强的企业获得了更高的持续收益增长。本文针对供应链风险管理主要进行了以下几点创新研究:1、提出了具有分形特征的供应链优化了信息空间的维数,优化了信息效率,研究了重构分形特征供应链的意义及其重构的理论与实践方法。2、基于分形理论研究了供应链风险结构及其特征参数。使供应链风险管理由两层延伸至更多层成为可能,拓展了供应链风险管理的视野,建立了供应链整体的风险管理观念。3、运用复杂理论的非线性模型对供应链风险扩散收敛机理进行了分析,研究了四种情形下的供应链风险扩散收敛模型。4、研究了分形供应链自相似性对收益的影响。供应链自相似性导致了供应链组成部分风险的自相关和行为的自组织,运用模型分析了这种自相似性对生产商、供应商和供应链整体均衡及收益带来的影响。更多还原

【Abstract】 Since the end of the last century, Supply chain management has become an important access to achieving competitive advantage and enhancing competitiveness,Network has become an integral part of the strategy method for enterprises responding faster to changing market. Network members become partnership and rely on each other by cooperating closely and integrating inseparably. Reducing transaction costs, focusing on core competencies, accessing to key technologies and sharing the risk have become the key to success for supply chain management. Supply chain takes competitive advantage to enterprises at the same time, it also brings potential risks. People increasingly concerned about supply chain risk, when they researched on supply chain network design, supply chain coordination and contract, supply chain inventory issues, supply chain performance measurement and supply chain information flow and value.Although we have an extensive research on the supply chain risk management, the literature rarely studies the supply chain as a whole using non-linear complexity theory. This paper from the perspective of fractal theory study the supply chain structure of the overall risk and risk management issues, using nonlinear analysis tools. As some initial exploration, the paper work and the main conclusion of the study are as follows:1. It elaborated the supply network vulnerability, supply network elasticity, supply network robustness, supply network flexibility, supply network reliability, supply chain agility, supply network architecture, supply chain complexity and so on relative concepts, analyzed their relationship with supply chain risk. It systematically summarized the domestic and foreign literature about supply chain risk management and carried on the brief summary to the present research.2. It proposed the fractal characteristic supply chain optimized the information space dimension, optimized the information efficiency. Through the analysis to the universality of supply chain risk factors and the complexity of the participants, it pointed out the huge information need for supply chain risk management. In order to whole effective control the supply chain risk, it must seek the effective information mechanism to enhance the information efficiency and achieve active control and management with the smallest information need.Analyzing on Tian Guoqiang’s decentralized information mechanism and Tian theorems, it obtained that the economic environment information space possesses fractal characteristics, which has the most superior information efficiency and the most superior resources deployment. So unifying the information mechanism and fractal theory analysis, it proposed to restructure supply network and establish fractal characteristic supply chain risk structure, then improved information efficiency of supply chain risk management.3. The paper highlighted the significance of restructuring fractal characteristic supply chain, and studied the restructuring theory and practice method. Basing on fractal theory, it provided suggestion to compress complex enterprise and supply chain into fractals with self-similarity schemas, so reduced the information dimension and decreased supply chain risk, people can manage and control more large-scale enterprise and longer supply chain only using lesser information. First, from the perspectives of competition efficiency, motivating mechanism and self-organization theory, we studied the significance of the restructuring fractal characteristic supply chain. Self-organization, self-adaptation and self-optimization results from self-similarity, it enhances supply chain efficiency and competency. Then the paper discussed practice and theory restructuring method for the supply chain risk structure. Theoretically , by the principal components analytic method it mainly selected key irrelative aspects of outstanding individual, therefore restructured supply chain with self-similarity characteristic according to such key aspects as standard core schemas. Practically, it restructured supply chain core enterprise with fractal characteristic according to six configurable aspects, such as organization structure, policy and procedure, monitoring system, enterprise culture, employee development system and information system. The paper proposed that the application of normalized schemas are expanded and penetrated to core enterprise interior and all kinds of suppliers and supplier’s suppliers, it suggested that individual organizations operate by similar mechanism, select and manage suppliers by similar mechanism, cooperate with multi-layer partner by similar mechanism, manage interior units by similar mechanism, hence we compress core enterprise and supply chain into fractals with self-similarity.4. Supply chain risk structure and its characteristic parameters were focus of this dissertation. It researched supply chain nested structure based on value activity, abstracted supply chain risk factors pursuant to the nested fractal structure, provided the framework of supply chain risk index system. It discussed the analytic theory for researching supply chain risk structure, and pointed out that fractal theory is the good tools to feature the supply chain risk structure. In the view of fractal theory, the paper described the calculation method of fractal dimension of value activity risk structure and of partial fractal dimension of multi-layer supply chain risk structure. It also studied the singularity spectrum of multifractals and characteristic parameters of supply chain risk structure. Multifractal spectrum uncovered diversity, irregularity, nonuniformity and complexity of risks from supply chain different area. Moreover, by changing fractal scale, people can analyze and manage supply chain risk structure from different scale and different scope. Finally, it gave an example to calculate the value of supply chain fractal dimension and characteristic parameters of supply chain multifractals. This article extended the supply chain risk management from two-layer to multi-layer, expanded visual field of supply chain risk management, refined new theory relevant to real-world supply chain risk, and established framework of supply chain risk management in whole view.5. This article has established models to simulate the diffusion and convergence of supply chain risk, and studied the management mechanism. Because of supply chain complexity and non-linear interrelation of supply chain core enterprise interior and between the supply chain level, we utilized the non-linear model of complex theory to research the process of diffusion and convergence of supply chain risk. Four kinds of models have been founded, when there are no interference of supply chain members, or there are, and when there are interference with time lag but no parameters disturbance, or when there are parameter disturbance but no interference time lag. These models emphatically analyzed the supply chain risk factors how to exist in the supply network and what are their styles of coming into being, conducting, diffusing, shaking, converging, and the stable state movement way. The essential model parameters include the diffusion and convergence parameter of supply chain risk, the average responsive time and the parameter of risk managing and controlling capacity.In view of the three essential model parameters, the article has discussed how supply chain risk structure affects the diffusion and convergence parameter of supply chain risk, and suggested the method of reducing supply chain complexity to decrease supply chain risk. It has discussed how supply chain agility influences the average responsive time of supply chain, and analyzed the evaluation model of supply chain agility and the method of improving supply chain agility. Eventually, it had discussed how performance measurement mechanism works on the parameter of risk managing and controlling capacity, and proposed the method of making good use of performance measurement mechanism to improving willingness of risk managing and controlling.We have discussed that supply chain with self-similarity can transmit accurate information in good time, and adapt easily to quick changing environment by self-organization and self-optimization, so the fractal characteristic supply chain can improve diffusion and convergence parameter of supply chain risk and reduce average responsive time, at the same time, mitigate the risks with small structure changes. The self-similarity of supply chain is propitious to compare performance of members, to select suitable performance evaluation mechanism. So the fractal characteristic supply chain can improve the three parameters of above-mentioned model, can enhance the ability of supply chain to avoid risks, mitigate risks, and eliminate risks.6. In this article, we have studied how the self-similarity of the fractal supply chain influences the income of supply chain members. It proved that the normalized schemas and self-similar structure induce strong pertinence relation between supply chain interior because of restructuring supply network based on fractal theory. Nevertheless, self-similarity of supply chain has two-sided function to risks, on the one hand, the self-similarity causes self-relation and enhances risk, on the other hand, self-similarity causes self-optimization and decreases risk, it is the two powers to affect risk evolution.The dissertation referred to Babich(2003) supplier credit risk correlation model and Martin A. Lariviere & Evan L. Porteus(2001) price-only contracts newsvendor model. The synthesis influence on the income of supply chain members has been researched from both self-relation and self-optimization functions resulting from self-similarity of supply chain. The paper disclosed the synthesis self-similarity influence separately from longitudinal and latitudinal direction in the supply chain, when manufacture and solo supplier are longitudinally similar and when manufacture and multi suppliers are latitudinally similar. Moreover, it analyzed the compositive result on the income of supply chain members under deterministic and nondeterministic demand.With one supplier and under uncertainty demand, the optimum equilibrium order quantity and profit of the integrated supply chain increase while product default risk decreases, the optimum equilibrium order quantity, supplier profit and manufacture profit of the non-integrated supply chain all increase while product default risk decreases. Because self-similarity of supply chain causes self-optimization and decreases the product default probability, so self-similarity of supply chain is useful to improve the income of both integrated and non-integrated supply chain.With two suppliers and under certainty demand, when the effect of self-optimization caused by self-similarity is greater to some extent, the optimum profit of manufacture and the supply chain increase while self-similarity improves. The effect on suppliers’profit is uncertainty, the profit of suppliers increase only when supplier product default probability is very high.With two suppliers and under uncertainty demand, by analyzing examples with exponential distribution demand curve or normal distribution demand curve, and with symmetrical suppliers, the optimum profit of manufacture and the whole supply chain increase while self-similarity of supply chain improves and when the effect of self-optimization caused by self-similarity achieves certain level, nevertheless, the optimum profit of suppliers drop.Because the profit of the whole supply chain goes up when self-similarity improves, supply chain coordination can inspirit suppliers.7. This article made case study about supply chain risk structure and management mechanism with fractal characteristic. First, it carried on the brief summary to development situation of franchised enterprises with higher self-similarity. Their good trend demonstrates that enterprises and supply chains with fractal characteristic can adapt fiercely changing and intensively competitive business environment, simultaneously control and keep away risks more easily and develop well. Then it analyzed two domestic representative snack enterprises which one operates Chinese food and the other operates west food, elaborated their structure features and management principles, calculated the fractal dimension of supply chain value activity, analyzed the singularity index and the singularity spectrum, depicted their difference of supply chain risk structure, calculated the relative value of their diffusion and convergence parameter of supply chain risk. With the thorough analysis of 2002 and 2007 year in several significant crisis events, the actual operation data of two enterprises proved that the more greater value the parameter of risk managing and controlling capacity is, the more lower risk the supply chain moves stable in. Though the relative parameter value of diffusion and convergence is bigger, so long as the relative average responsive time is short and respond quickly, risk crisis of supply chain can be effectively managed and come back to stable state. Through comparing many years actual operation data of enterprises, the result is found that the higher self-similarity enterprise grow up persistently and get higher income growth. Although higher self-similarity cause higher self-relation, it also produces strong capacity of self-organization, self-optimization, mitigating and keeping away risks.This article has mainly conducted following several innovation researches in view of supply chain risk management:1. It proposed that fractal characteristic supply chain optimizes the information space dimension, optimizes the information efficiency. Then it highlighted the significance of restructuring fractal characteristic supply chain, studied the restructuring theory and practice method.2. Based on fractal theory, it researched supply chain risk structure and its characteristic parameters. This article extended the supply chain risk management from two-layer to multi-layer, expanded visual field of supply chain risk management, and established framework of supply chain risk management in whole view.3. This article has established non-linear models to simulate the diffusion and convergence of supply chain risk according to complex theory, and studied models under four situation.4. The paper studied how the self-similarity of the fractal supply chain influences the income of supply chain members. The self-similarity of supply chain causes self-relation of supply chain risk, and causes self-optimization of supply chain. It established model to analyze the influence on income and equilibrium of manufacture, suppliers and the whole supply chain.更多还原