| A supply chain is a network of facilities and distribution options that performs the functions of procuring raw materials, transforming them to products and distributing the finished products to the customers. The modern supply chain is a highly interconnected network of facilities that are spread over multiple locations and handle multiple products. In a highly competitive global environment, optimal day-to-day operations of supply chains is essential.;To facilitate optimal operations in supply chains, we propose the use of Model Predictive Control (MPC) for supply chains. We develop: 1. A new cooperative MPC algorithm that can stabilize any centralized stabilizable system 2. A new algorithm for robust cooperative MPC 3. A state space model for the chemical production scheduling problem;We use the new tools and algorithms to design model predictive controllers for supply chain models. We demonstrate: 1. Cooperative control for supply chains: In cooperative MPC, each node makes its decisions by considering the effects of their decisions on the entire supply chain. We show that the cooperative controller can perform better than the noncooperative and decentralized controller and can reduce the bullwhip effect in the supply chain. 2. Centralized economic control: We propose a new multiobjective stage cost that captures both the economics and risk at a node, using a weighted sum of an economic stage cost and a tracking stage cost. We use Economic MPC theory (Amrit, Rawlings and Angeli, 2011) to design closed-loop stable controllers for the supply chain. 3. Integrated supply chain: We show an example of integrating inventory control with production scheduling using the tools developed in this thesis. We develop simple terminal conditions to show recursive feasibility of such integrated control schemes. |
