Comparison of Linear Quadratic Regulator and Model Predictive Control Based Algorithms in Continuous Production

• Published in: Bulletin of the Polytechnic Institute of Iași. Electrical Engineering, Power Engineering, Electronics Section Vol. 69; no. 2; pp. 9 - 34
• Main Authors: Chindruş, Amelia, Copoţ, Dana, Căruntu, Constantin-Florin
• Format: Journal Article
• Language: English
• Published: Iasi Sciendo 01.06.2023; De Gruyter Brill Sp. z o.o., Paradigm Publishing Services
• Subjects: continuous manufacturing; Cyber-physical system; Industry 4.0; linear quadratic regulator; model predictive control
• ISSN: 2537-2726; 1223-8139; 2537-2726
• DOI: 10.2478/bipie-2023-0007

The integration of Industry 4.0 into manufacturing processes necessitates the automation of complex, large-scale operations within cyber-physical systems (CPSs). Pharmaceutical manufacturing, in particular, requires a transition from traditional batch processing to continuous manufacturing to achieve seamless integration with CPSs. This paper explores the comparison between two control strategies for pharmaceutical tablet production: the linear quadratic regulator (LQR) method and an established model predictive control (MPC) algorithm. The LQR method focuses on providing optimal stability and robustness for the plant’s operations, particularly through centralized management of key process units in the dry granulation process. A detailed plant model is utilized to test the performance of the LQR controller, with results benchmarked against those obtained using the MPC algorithm.