【作者】 罗卫；

【导师】 张子刚； 欧阳明德；

【作者基本信息】 华中科技大学 ， 管理科学与工程， 2005， 博士

【摘要】 本文对一般供应链中的牛鞭效应和库存方差进行了研究,着重讨论和分析了牛鞭问题对供应链的影响,研究是基于一个普遍的生产计划算法,它可以称为自动渠道的、基于库存和定购的生产控制系统(APIOBPCS,Automated Pipeline,Inventory and Order Based Production Control System )。对研究目的和意义以及APIOBPCS 策略进行了简要介绍,对牛鞭效应的类型和产生的根源做了简要说明,对相关研究领域和研究现状进行了较详细的回顾,对目前学术文献中有关供应链牛鞭效应的主要研究理论、方法、测量标准和结论做了简要评述。使用因果循环图、框图、差分方程和z-变换对APIOBPCS 供应链系统建立了一个传递函数模型,依据该模型,对APIOBPCS 供应链系统的稳定性和鲁棒性进行了讨论,通过分析发现非稳定性是由于低劣的供应链设计方案引起,在供应链内部针对专门的生产延迟,选择适当的参数,可以使两个反馈环达到和谐并且避免供应链系统的不稳定性。这种方法也可以推广用在其它具有生产延迟和分销环节的供应链中。从控制工程的角度,对APIOBPCS的重要变体DE-APIOBPCS供应链产生的牛鞭效应,推导出了其分析表达式。同时还推导出库存水平方差的分析表达式,通过把它与牛鞭效应分析表达式一起使用的做法,对两种方差之间一系列权重进行了讨论,由此确定一些恰当的定购系统设计方案。对定购适当(OUT,order-up-to )补充策略与不同需求预测方法一起应用时所引起的牛鞭效应进行了讨论,指出一般性生产控制策略,即自动渠道的、基于库存和定购的生产控制系统(APIOBPCS)策略可以避免定购方差放大,在供应链中产生了平滑定购类型。讨论了多阶段供应链中信息共享的影响; 当使用OUT 策略时,由传统供应链和信息共享供应链的比较,发现后者产生的牛鞭效应(定购数量的方差放大)减少很多,尤其是在供应链的较高阶段,牛鞭效应减少得更明显,但是,定购数量仍然会沿着供应链向上传递时不断增大。当使用平滑APIOBPCS 策略时,把传统供应链和更多还原

【Abstract】 This dissertation investigates Bullwhip Effect and inventory variation generated in a general supply chain and emphasizes the impact of the Bullwhip on the supply chain and considers a well-established production-scheduling algorithm, called the Automatic Pipeline, Inventory and Order Based Production Control System (APIOBPCS). The introductions about the aim and implication of this research and APIOBPCS policy are made, and the types and the induced sources of the Bullwhip effect simply explained. Relevant research fields and status quo are reviewed in detail, and the main research theories, methods, measures and conclusions related to the Bullwhip effect generated in the present scholastic literatures are simply demonstrated. A transfer function model of the system through using causal loop diagrams, block diagrams, difference equations and z-transforms is built up and the stability and robustness of the APIOBPCS supply chain system via this model is discussed. It is shown that instability arising from poor design and its avoidance via the appropriate parameter settings for tuning the two feedback loops within the supply chain for a specific production delay. The procedure is readily extended for other production delays and distri-butions. From the control engineering perspective, when an analytical expression for bullwhip generated in the supply chain based on the special case of APIOBPCS model, i.e. the DE-APIOBPCS model is derived, an analytical expression for the variance of the inventory position is also derived and used together with the bullwhip expression to determine suitable ordering system designs that minimize both bullwhip and inventory variance for a range of weightings between the two variances. When order-up-to replenishment policies are used with the different forecasting methods, the bullwhip effect generated in simple supply chain is analyzed. Whereas a generic production control replenishment policy called APIOBPCS is demonstrated to avoid demand variance amplification and generates smooth ordering patterns in the supply chain. The beneficial impact of information sharing in multi-stage supply chains is discussed. With the use of OUT policies, via comparing the traditional supply chain with the information shared supply chain, it is showed that the bullwhip effect (variance amplification of ordering quantities) is reduced significantly in the latter with the help of information sharing, especially at higher levels in the chain but variance of ordering quantities still increases as one moves up the chain. When using the smoothing APIOBPCS policies, it is showed that information sharing helps to reduce order variance at the higher levels of the supply chain. This dissertation compares the expected performance of a vendor managed inventory (VMI) supply chain with a traditional "serially linked" supply chain both based on APIOBPCS. When adopting VMI, it is showed that the rationing and gaming or the Houlihan effect, and the order batching effect or the Burbidge effect may be completely eliminated, however, Forrester effect is less clear cut. VMI is significantly better at responding to volatile changes in demand such as those due to discounted ordering or price variations etc.更多还原