The Study On Single-period Inventory Management Under Uncertain Supply And Demand

Inventory management is the management and control of all kinds of goods,finished goods and other resources in the whole process of manufacturing and ser-vice industry.The purpose of inventory management is to keep the reserves at a reasonable level.In the process of inventory management,there are many uncertain factors,such as uncertain supply and uncertain demand.In real life,those uncer-tainties are highly difficult to be controlled.When the market demand can not be met in time,the enterprise will bear the shortage loss.Therefore,in order not to lose more customers,the enterprises can prepare reserve stock in advance to reduce the loss caused by uncertain factors.At the same time,it will increase the cost,such as the cost of inventory or the disposal cost of surplus products.In order to solve the problem of inventory management with uncertainties,this dissertation first studies the single-stage inventory management problem under un-certain supply,where the uncertain proportion of supply from primary supplier is characterized by a fuzzy variable.Based on credibility optimization theory,we build the credibility equivalent value convex programming model.Under the triangular and trapezoidal supply modes,the obtained models are analyzed and the feasible domain decomposition method is summarized.And finally the global optimal solu-tion is got by comparing the local optimal solutions.Furthermore,this dissertation studies the problem of multi-stage inventory management under uncertain supply and demand.Because of the retailer's multiple order opportunities,we formulate a credibilistic dynamic programming model.The uncertainties of supply and demand are characterized by the credibility distribution.We use the expected value and equivalent value as measure indexes,and get the mean convex programming model.Based on the credibilistic dynamic programming model,the mean cost of each stage is analyzed,and the best order quantities are worked out.The main contribution of this dissertation can be summarized as follows:(?)In fuzzy environment,the uncertainties are described by credibility distributions,and the credibility equivalent value model is established;(?)We analyze and discuss the property of the credibility equivalent value model,and generalize the solution method based on the feasible domain decomposition;(iii)In order to solve the prob-lem of multi-stage inventory management,the credibilistic dynamic programming model is formulated;(iv)We analyze and discuss the property of the credibilistic dy-namic programming model,the general expression of the optimal order quantity for the nth stage is derived.(v)Some numerical experiments are provided to illustrate the modeling ideas and the effectiveness of the solving methods.The sensitivity analysis of each parameter is given.