Comparative Analysis Of Multi-period Non-stationary Inventory Control Models

With the progress of science and technology and the development of productivity,customers' demand is more and more personalized,and the life of product cycle is continuously shortening,additional some other uncertainty factors,making randomness and instability of a lot of products.All those brought great challenge to enterprises' efficient production and inventory management.Therefore it is both of significant and necessary to study the problem of inventory control under nonstationary demand.This thesis studied multi-period inventory control problem with stochastic nonstationary demand,order lead time and service level constraints.First of all,a general framework of inventory decision was established.Then,a stochastic inventory control problem which aiming to minimize the total inventory cost using reorder point and order quantity as decision variables was studied.A static model and a static-dynamic model were given respectively,and the corresponding solution methods were given.The deficiency of the existing model was discussed and a remedy to it was given too.Then different forecasting models were used to predict the future demand and inventory control decision were given with several different demand patterns.Simulation studies were done to test the performance of the two control models.The simulation results showed that the performance of the two models is different under different demand patterns.In the case of cyclical demand,the difference is not so significant,while there are obvious difference in the case both of trend demand and mixed demand patterns.Sensitivity analysis shows that the more fluctuation in demand,the more difficult it is to control the inventory.Those results provide some suggestion to choose different control method under different demand patterns.