On November 22, we organized a scientific seminar on “Multi-Fidelity Modeling for Enhanced Process Monitoring“. The scientific seminar was led by Rastislav Fáber, a PhD student at STUBA. The scientific seminar was organized in the framework of the FrontSeat project as part of the seminar series on “Research Seminar on Smart Cybernetics”.
Abstract:
Effective process monitoring is critical for optimizing refinery operations. This study investigates a multi-fidelity (MF) modeling approach, combining dynamic low-fidelity (LF) data from online analyzers with high-fidelity (HF) laboratory samples taken infrequently. Using a comprehensive industrial dataset, we compare static and dynamic models in an MF framework to improve decision-making. A Gaussian process (GP)-based correction is applied to align LF model outputs with HF measurements. While static models provide computational efficiency, dynamic models more effectively capture time-dependent behaviors, achieving higher accuracy. Our results highlight the potential of dynamic MF modeling for improved process monitoring and control.
This project has received funding from the European Union’s Horizon under grant no. 101079342 (Fostering Opportunities Towards Slovak Excellence in Advanced Control for Smart Industries).
0 Comments