Computational results for the control of a divergent N-echelon inventory system

Consider a divergent multi-echelon inventory system, e.g., a distribution system or a production system. At every facility in the system orders are placed (or production is initiated) periodically. The order arrives after a fixed lead time. At the end of each period linear costs are incurred at each facility for holding inventory. Also, linear penalty costs are incurred at the most downstream facilities for backorders. The objective is to minimize the expected holding and penalty costs per period. Within a class of practically useful policies the decomposition result is used to develop an algorithm which determines the control parameters of a near cost-optimal replenishment policy. A simulation study of a divergent three-echelon system reveals that this algorithm performs well.