Determining Process Target Value and Ordering Policies in Two-Echelon Hierarchy Production System.

Abstract

Selection of the mean (target value) for a production process is a classical problem in quality control. Basically, a process mean is selected based on a balance between production cost and economical consequences associated with conforming items and nonconforming items. The process mean affects many production decisions. In particular, because it determines the process conforming and yields rates, it affects the production setup policy. These production decisions also affect the raw material requirement and, thus, its procurement policy when the raw material is supplied by outside vendors. Consequently, process mean, production and raw material procurement policies should be jointly determined to minimize the total related costs. Furthermore, in practice, quantity discounts may be available in raw material purchasing. Because of the interaction between the process mean determination and the raw material ordering policy, quantity discounts will affect all of the related decisions. This dissertation consists of three parts. In the fist part, a two-echelon model is formulated to incorporate the issues associated with production setup and raw material procurement into the classical process mean problem for a single-product production process. In the second part, quantity discounts in raw material purchasing are incorporated into the model. The quantity discounts policy under study is known as all-unit quantity discounts. In the third part, we consider a situation in which the supply rate of the raw material is finite and constant. Three cases in terms of quantity discounts in the raw material purchasing are considered: no discounts, incremental quantity discounts and all-unit quantity discounts. Mathematical models are formulated for all the cases discussed above. Analytical properties are derived and efficient solution algorithms are proposed. Examples are used to illustrate the solution procedures and sensitivity analyses are performed to study the effects of model parameters on the optimal solutions.