Safe Model-Based Design of Experiments using Gaussian Processes
Abstract
Construction of kinetic models has become an indispensable step in the development and scale up of processes in industry. Model based design of experiments (MBDoE) has been widely used for the purposes of improving parameter precision in highly nonlinear dynamic systems. This process needs to account for both parametric and structural uncertainty, as the feasibility constraints may well turn out to be violated leading to a unsafe experimental condition when an optimally designed experiment is performed, leading in the best case to less informative data sets. In this work, a Gaussian process is utilized in a twofold manner: 1) to quantify the uncertainty realization of the physical system to calculate the plant-model mismatch, 2) to propagate the parametric uncertainty. This method provides a guarantee for the probabilistic satisfaction of the constraints in the context of model-based design of experiments. The method is assisted with the use of adaptive trust-regions in order to facilitate a satisfactory local approximation. The proposed method is able to help the design of optimal experiments starting from limited preliminary knowledge of the parameter set, leading to a safe exploration of the parameter space. The performance of this method is demonstrated through illustrative case studies regarding the parameter identification of kinetic model in flow reactors.
Instructions to log on
You are invited to a Zoom webinar.
When: Oct 1, 2020 13:00 PM London
Topic: CPSE Summer Webinar Series: Dr Panagiotis Petsagkourakis (CPSE, University College London)
Please click the link below to join the webinar:
https://zoom.us/j/91341312739?pwd=dkU2VFBacktRQ2pMTUVhK1JneFpRdz09
Passcode: 7Jd8Nk
Or iPhone one-tap :
US: +19292056099,,91341312739#,,,,,,0#,,529128# or +12532158782,,91341312739#,,,,,,0#,,529128#
Or Telephone:
Dial(for higher quality, dial a number based on your current location):
US: +1 929 205 6099 or +1 253 215 8782 or +1 301 715 8592 or +1 312 626 6799 or +1 346 248 7799 or +1 669 900 6833
Webinar ID: 913 4131 2739
Passcode: 529128
International numbers available: https://zoom.us/u/acNI3XCayA
About CPSE
The Centre for Process Systems Engineering (CPSE) is a multi- institutional research centre. It was inaugurated in August 1989 by Professor Roger W.H. Sargent and involves Imperial College London and University College London. Innovative and dynamic in its approach to research and development, CPSE’s accomplishments include the Queens Prize for Higher Education, presented in 2003 for research excellence and technology transfer. Three spin-out companies have been established. One of these, Process Systems Enterprise Limited (PSE), received the Royal Academy of Engineering’s MacRobert Award in 2007; this is the UK’s highest award for innovation in engineering. The Centre continues to broaden the scope of its activities over all size scales enabling all areas of process systems to be addressed. This enables CPSE to be responsive to the evolving needs of industry and society.