Application Research On Improved Sequential Bifurcation Method In Multi-echelon Inventory System

With the deepening of the global integration process,the enterprise system module and supply chain model are becoming more and more complex,the target of the enterprise is becoming more and more diverse,so the factors of the simulation model is much more than before.The target of the simulation model is no longer a single,but multiple.so the simulation experiments will require a lot of time and cost,which is undoubtedly a serious waste.In this case,we need to continue to study the factors of screening methods,especially the multi-response sequential bifurcation screening method.If all factors are screened in the initial simulation to find out the important factors for the simulation model,it is easily to simplify the process in the next step of the simulation experiment.And it also reduces the time and the cost.Therefore,it is more and more important to research the method of screening factors.This paper describes the background and the significance of researching the screening factors in the simulation experiments,and summarizes different methods to screen factors form domestic and foreign scholars,then it introduces the basic theory and concepts which relate to the method used in this paper.The third chapter describes the concept model and computer simulation model of the multi-echelon inventory system,then according to the principle of grouping,the factors involved are grouped and the accuracy of computer simulation model is verified.The forth chapter introduces the sequential bifurcation screening methods,then analyzes and compares the method of the unbiased main-effect estimator in the sequential bifurcation screening methods,and it is proved that this method has no bias.Finally,this method is used in multi-echelon inventory system to find out the important factors for the two outputs of the experiment respectively.The sequential bifurcation method is improved in the next chapter so that the two outputs can screened at the same time,the new method is used in the same simulation model and identify the same important factors.So the new method is more efficient and accurately by comparing the screening results of two methods.The innovative points in this paper are summarized as follows:(1)This paper analyzes and compares the method of the unbiased main-effect estimator in the sequential bifurcation screening methods,and it is proved that this method has no bias.(2)In this paper,it is improved the sequential bifurcation screening method,so that the two outputs can be screened at the same time;and present a method for determining the final number of replicates by the initial number of replicates,namely the same screening accuracy can be obtained by using the less number of replicates.(3)The improved sequential bifurcation screening method is applied to the same multi-echelon inventory simulation model,then the results are compared,and the efficiency of the improved method is proved in the practical application.