Sensitivity Analysis for Base-Stock Levels in Multiechelon Production-Inventory Systems

摘要: Effective management of inventories in large-scale production and distribution systems requires methods for bringing model solutions closer to the complexities real systems. Motivated by this need, we develop simulation-based estimating sensitivities inventory costs with respect policy parameters. These sensitivity estimates are useful adjusting optimal parameters predicted a simplified that can be incorporated simulation. We consider capacitated, multiechelon operating under base-stock policies estimators derivatives levels. We show these converge correct value finite-horizon infinite-horizon discounted average cost criteria. Our easy implement experiments suggest they quickly. illustrate their use optimizing levels subsystem PC assembly system major computer manufacturer.