A myopic policy based simulation optimization framework in a general decentralized supply chain

Decentralized multi-echelon supply chains are often computationally intractable and difficult to analyze due to the complex network structure and various decision makings. This paper develops a simulation optimization framework for a general decentralized supply chain composed of sup-pliers, plants, distributors and customers. Each facility in the network is under local control and makes decisions based on a myopic policy. The simulation optimization framework is composed of an event simulator, an optimization engine, a system updater and a process controller. The simulation test bed can help the network facilities asymptotically optimize their decisions based on the feed-back during the simulation process, and evaluate the performance of each facility.