Application of Model Predictive Control to Robust Management of Multiechelon Demand Networks in Semiconductor Manufacturing

Abstract

Model predictive control (MPC) is presented as a robust, flexible decision framework for dynamically managing inventories and satisfying customer demand in demand networks. In this paper, a formulation and the benefits of an MPC-based, control-oriented tactical inventory management system meaningful to the semiconductor industry are presented via two significant examples. The translation of available information in the supply chain problem into MPC variables is demonstrated with a single-product, two-node supply chain example. Simulations demonstrating the ability of a properly tuned MPC control system to maintain performance and robustness despite plant-model mismatch are shown. Insights gained from these simulations are used to formulate a partially decentralized MPC implementation for a six-node, two-product, three-echelon demand network problem developed by Intel Corporation. These simulations show that the demand network is well managed under conditions that involve simultaneous demand forecast inaccuracies and plant-model mismatch.