”Model-based control – from industrial process optimization to new biomedical products”
Abstract:
Feedback and optimization enabled model-based control for dynamical systems is central to new products and processes.
A unified approach to model-based experimental design, state- and parameter-estimation, prediction error method based system identification, and predictive control along with novel advances in scientific computing for such systems enable significantly improved operation of oil fields, new processes such as the U-loop reactor for production of single-cell protein to provide food for a world with 9 billion people, the green transition by electrification of our energy system, and new health technologies such as the artificial pancreas for people with diabetes.
Model predictive control that is based on stochastic differential equations enable all these processes and products and we present these cases along with the underlying computational methods for modelbased control technology. The important role of high-performance and scientific computing is illustrated through the case studies.