【作者】 封云；

【导师】 马军海；

【作者基本信息】 天津大学 ， 管理科学与工程， 2008， 博士

【摘要】 二十世纪八十年代以来,随着全球经济一体化进程的不断加快,先进的管理理念和信息技术促进了企业生产方式和市场环境的转变,原有的企业与企业间价格、质量等简单个体因素的竞争变成了与企业与企业之间的供应链竞争。这种竞争把企业供应链管理提高到一个更高的高度,供应链管理己经成为世界性的重要研究课题。造成供应链效率低下的牛鞭效应作为供应链管理领域特有的现象,更多的引起了人们的关注和重视。对其进行系统的分析和综合研究具有重要的理论意义和实际意义。本文在国内外研究成果的基础上,对牛鞭效应问题进行了深入研究。分别从传统统计学方法、系统控制理论方法、混沌理论方法三个方面出发,完善供应链管理理论中关于牛鞭效应问题的研究。总结了传统统计学学领域AR(1)、ARMA(1,1)需求时间序列模型牛鞭效应大小的论证工作;并从系统动力学角度提出缓解牛鞭问题的O-S反馈控制方法,从而为能够消除牛鞭提供理论支持;此外,本论文还从复杂性的角度对采用混沌预测方法来预测和处理需求放大问题的工作做了初步的研究,是未来这一研究工作的很好铺垫。本论文的主要内容和创新点如下:(1).系统全面的总结了供应链管理中牛鞭效应问题的国内外研究现状,分析了目前牛鞭效应研究的三种主要研究方法(传统统计学方法、系统控制理论方法和混沌理论方法)的特点。在此基础上提出了牛鞭效应研究按照应用时间应该分为四个阶段。(2).基于传统统计学研究牛鞭效应方法,重点推导了移动平均预测需求方法下,AR(1),ARMA(1,1)时间序列模型的牛鞭效应的大小,证明了基于这两种时间序列模型,牛鞭效应始终存在,讨论了移动平均预测参数在不同取值下,对供应链牛鞭效应的影响。(3).详细介绍系统控制理论方法在研究牛鞭效应问题中的应用,提出采用O-S控制反馈方法可以表征牛鞭问题和供应链复杂网络结构,并通过O-S反馈控制参数可以减低牛鞭效应的大小,讨论了O-S反馈控制方法在供应链结构表征上的应用。(4).讨论了混沌理论方法在研究牛鞭问题上的应用,研究了N级供应链模型所组成的复杂动力学系统,计算了均方误差预测需求方法下,N级供应链模型Lyapunov指数的大小,讨论了基于混沌理论价格-需求模型的混沌特性,详细介绍了牛鞭效应非线性预测的不同方法,并通过实证比较了不同非线性需求预测方法的优劣。更多还原

【Abstract】 With the development of global economy integration process, the firms’manufacture and market mode have changed by the advanced management and technology. Firms have been gradually regarding the competing as part of a supply chain against other supply chains, rather than as a single firm against other individual firms since the 1980s. It brings much new hard work for the supply chain management Supply chain management has been the important subject all over the world. Interesting in the bullwhip effect, the peculiar phenomenon in supply chain management, has steadily increased when it is regard -ed as the major cause of supply chain ineffeiciency. It is important to anlyze the bullwhip effect systematically and comprehensively.Based on the literature research the approach on the bullwhip effect from three perspectives and analyzed: statistic methods, system control theory and chaotic theory. It assumed that the bullwhip effect based on the model of series time AR(1) or ARMA(1,1) and supposed a new methods, named O-S feedback control methods based on system dynamics theory, to support the study of bullwhip effect. Meanwhile, we also introduce the demand nonlinear forecasting methods, based on chaos theory, to decrease the supply chain dynamics. The corresponding control measures to the bullwhip effect are proposed in the last section. The main content and findings in this dissertation are summarized as follows:First, a general literature review relating to bullwhip effect in supply chain management is presented. We analyze the main three methods in the research of bullwhip effect in supply chain management: statistic methods, system control theory and chaotic theory. For each of them, a brief description of the basic content and character is provided. Next, based on the traditional statistic, we use the model of AR(1), ARMA(1,1) serial times to detrude the formula of bullwhip effect in N-echelon and prove the bullwhip effect is exist under the two series model. Moreover, we discusses the influence of model’s parameters.Thirdly, we summarize the theory of system control in the research of bullwhip effect and suppose new methods, named O-S feedback control theory, to describe the supply chain net structure. Through controlling the O-S feedback parameter, the influence of bullwhip effect can be alleviated.Fourthly, we systematic introduce chaos theory as the new method to reduce bullwhip effect. We take N-echelon supply chain model as dynamic system and calculate maximal Lyapunov exponent . The result shows the model is chaotic. In the last, the different nonlinear method of demand’s forecasting was compare in order to get the accurate forcasting result.更多还原