Inventory Control and Demand DistributionCharacterization

Inventory control;Supply chain

Bai, Liwei ; Industrial and Systems Engineering

University:Georgia Institute of Technology

Department:Industrial and Systems Engineering

关键词: Inventory control;    Supply chain;   

Others  :  https://smartech.gatech.edu/bitstream/1853/6847/1/bai_liwei_200505_phd.pdf

美国|英语

来源: SMARTech Repository

【 摘 要 】

This thesis studies three problems related to inventory control.The first problem is motivated by the need to eliminate thebullwhip effect in a supply chain. An important source of thiseffect is the inventory control policy, which is originallydesigned to smooth production in response to demand variationalong the supply chain arising from the customers. To address thisissue, we propose an estimation method based on the controlvariate technique. A byproduct of this approach is a stabilizinginventory control policy. We evaluate the effectiveness of theproposed method using the models from the literature.Generally, the derivation of the inventory policies requires theknowledge of the specific demand distribution. Unfortunately, inseveral cases the demand is not observable in a direct way. Thesecond problem is motivated by a practical application where onlypartial demand information is observable. Towards this end wederive estimators of the first two moments of the (daily) demandby means of the renewal theoretical concepts. We also propose aregression-based approximation to improve the quality of theestimators. A series of numerical studies are carried out toevaluate the accuracy and precision of the estimators and toinvestigate the impact of the estimation on the optimality of theinventory policies.The last part of this dissertation studies a periodic-reviewinventory system with regular and emergency orders. Emergencyorders, characterized by shorter lead-time, higher ordering costand higher setup cost, are placed when the inventory level becomescritically low. Based on our assumptions, we formulate a dynamicprogramming model and prove the optimality of state-dependent sStype polices for both emergency and regular orders. We also deriveanalytic properties of the optimal policies. We gain somemanagerial insights into the optimal policies and cost performancefrom numerical studies.