A Forecasting-based Distributed MPC Approach on Inventory Level Control of Supply Chain Network

The development of economic globalization has led to an increasing complexity in supply chain networks(SCNs). However, the inventory control of SCNs has not been sufficiently investigated. In this paper, to maintain steady inventory levels for all entities in the SCN in the presence of demand fluctuations and uncertain lead times, we first propose a more realistic SCN model, within which each entity may have multiple providers and the in-transit stocks are excluded from the real inventory. Second, to mitigate the bullwhip effect, we employ a generic demand forecasting mechanism that only needs the historical information of consumers' demands. Finally, we introduce a forecasting-based distributed model predictive control (FDMPC) algorithm to keep the inventory levels of all entities steady and reduce the bullwhip effect. Numerical simulations confirm that the FDMPC approach effectively mitigates the bullwhip effect along the supply chain.