Analyze of demand and lead times sensitive multi echelon inventory decisions in supply chains and an industrial application

There are many approaches that are proposed in literature about multi-echelon inventory management in supply chains. It is determined by comparative analysis of existing approaches that, there is a necessity of new models and methodologies currently that support the inventory management process from removal of demand and lead time uncertainties to computation of cost and performance analysis. In this thesis study, titled ?Analysis of Demand and Lead Time Sensitive Multi Echelon Inventory Decisions and an Industrial Application?, an integrated methodology that covers all phases of multi-echelon inventory management process and deterministic/stochastic-fuzzy-neural models within the context of this methodology are presented. In the presented methodology, the variables that the models need are determined by neuro-fuzzy calculations, several assumptions and/or experts, and these values are combined in the developed cost model. The methodology contains the performance analysis of the model and that differentiates it from previous studies. The most important contribution that the proposed model makes to the multi-echelon inventory management literature is to present systematic deterministic and stochastic models, to determine demand and lead time variables by approaching them more realistic view and to calculate the chain cost. These models are for multi-echelon supply chains with tree structure and use less complex calculations to reach more proper and realistic results, and the deterministic model gets into the stochastic model structure. The presented methodology and models are used in an application step by step in a threeechelon food supply chain. Then satisfactory results are gained, and it is pointed out that the depot and chain performances are got better. Thus, the applicability of the methodology and models is validated with a real application. Keywords: Supply chain management, multi-echelon inventory management, artificial neural networks, neural-fuzzy networks, SCOR model.