Towards a "Turnkey" Model Predictive Controller

Steven Kuntz

For three decades, model predictive control (MPC) has been the flagship advanced control method in the chemical process industries. However, most implementations still use heuristic methods for designing MPC estimators, especially for offset-free MPC implementations.

I am working on a “turnkey” model predictive control (MPC) algorithm, that is, a model-based controller that works out-of-the-box, with no tuning necessary. I am developing system identification algorithms for integrating disturbance models, and have validated them in real-world applications [1]--[4]. Currently, I am implementing state-of-the-art optimization algorithms for fitting linear dynamical system models and developing a new stability analysis method for evaluating MPC with plant-model mismatch.

References:

[1] S. J. Kuntz and J. B. Rawlings, “Maximum likelihood estimation of linear disturbance models for offset-free model predictive control,” in American Control Conference, Atlanta, GA, 2022, pp. 3961–3966. doi: https://doi.org/10.23919/ACC53348.2022.9867344.

[2] S. J. Kuntz, J. J. Downs, S. M. Miller, and J. B. Rawlings, An industrial case study on the combined identification and offset-free model predictive control of a chemical process, FOCAPO/CPC 2023, January 8-12, 2023, San Antonio, Texas, 2023.

[3] S. J. Kuntz, J. J. Downs, S. M. Miller, and J. B. Rawlings, “An industrial case study on the combined identification and offset-free control of a chemical process,” Computers & Chemical Engineering, vol. 179, p. 108 429, 2023. doi: https://doi.org/10.1016/j.compchemeng.2023.108429.

[4] S. J. Kuntz, J. J. Downs, S. M. Miller, and J. B. Rawlings, An industrial case study on the combined identification and offset-free control of a chemical process, AIChE Annual Meeting, Orlando, FL, 2023. url: https://aiche.confex.com/aiche/2023/meetingapp.cgi/Paper/674827.