Development and implementation of a framework for Model Predictive Control of discrete-event manufacturing systems

Abstract This paper presents an open-source framework for Model Predictive Control (MPC) tailored for discrete-event manufacturing systems. Traditional MPC requires extensive mathematical modeling, which limits flexibility and adaptability. This research introduces a topology-driven approach where MPC controllers are synthesized automatically based on plant structure rather than on detailed equations. The framework, designed in Python and utilizing Google’s CP-SAT solver, addresses the automatic synthesis of an MPC controller based on the latest available information on the plant’s layout and state. It is adaptive and reconfigurable, meaning that it self-adapts to changes in the plant without the need of manual configuration. This topology-driven MPC offers a scalable and cost-effective solution to meet the demands of dynamic manufacturing environments.