A centralized optimization of dairy supply chain based on model predictive control strategy

This paper describes a simulation method based on model predictive control strategy to find the optimal decision variables to maximize profit in dairy supply chains with multi-product, multi-echelon distribution networks and multi-product batch plants. The key features of this paper are: 1) inducts a model predictive control strategy under centralized management approach which is suitable for dairy supply chains founded by the key manufacture; 2) proposes a feasible and convenient simulation method with GAMS simulation tool that is able to solve the multiperiod MILP optimization problem and thus provide a strong support to supply chain managers in the decision-making process. A performance comparison between model predictive control strategy and common static optimization is made to prove the excellent effect of MPC strategy with both profits and customer service level more than doubled in an example problem, then the paper compares and analyzes the behavior of supply chains under centralized and decentralized management approaches and shows the former yields better overall performance in dairy supply chain management