Research On Synergy Prediction Of VMI On Discount Strategy

VMI，which is short for Vendor Managed Inventory, has broken the traditionalfragmented way of inventory management and is considered to be the implementation ofintegrated supply chain. However, as we dig deeper into the VMI model and applied it towidespread use, some of the defects and limitations are also gradually exposed. Due tothe fact that this model is a supplier one-way replenishment decision-making process,namely the information used by suppliers to make market forecast only based onretailer’s historical inventory information and POS data, the result only reflects marketdemand which can be predicted according to past information. But the retailer’s discountplan is not taken into consideration. In fact, the consumer demand will be changedsignificantly during and after the promotion. Thus, deviation can be observed when wecompare the actual consumer demand with the predicted result of VMI model; this mayeven result into large fluctuations happening to the whole system.Under the circumstances that traditional VMI model suffers from low predictionaccuracy, limited information feedback speed and lack of capacity in inventory control,this thesis leveraged the principle of collaborative prediction in CPFR and designed VMIcollaborative prediction model. This model is supposed to overcome the disadvantagesand defect in traditional VMI model, help to produce more accurate demand predictionfor enterprises in supply chain, and control the inventory violate to improve the value forsupply chain.Firstly, the limitation in VMI model is summarized and the concept for collaborativeprediction is introduced, giving birth to VMI collaborative prediction model. Then, theoperating mechanism of this model is studied and the VMI collaborative prediction isdivided into three sub systems. At last, we modeling and simulate the computation fortraditional and collaborative prediction model of VMI using systematic dynamics. Wejustify the rational of collaborative prediction model by observing and comparing theinventory level of enterprises in supply chain node and the whole system, the inventoryviolation during promotion produced by two models during promotion. In addition, theconclusion has been made for the analysis result on inventory level and factors incollaborative prediction (prediction early days and safety inventory coefficient).