Webinar: Bayesian Adaptive Design with Chemical Processes

• Liam Fleming (Newcastle University)

Chemical processes are subject to considerable uncertainty arising from inherent stochasticity and process noise. On-plant experimentation is costly and time-consuming so approximating “digital twins” are constructed using simulation. A designed experiment is the starting point to providing responses from which to build a model to emulate the process. Gaussian Process (GP) regression provides a highly flexible model class to capture non-linearities in the process data and develop a predictive model. The webinar describes how Bayesian sequential design is used to maximise information gain for each experimental trial. A process simulator, modelled in Aspen Plus is used as a surrogate for a real chemical process, to demonstrate the capabilities of the methodology.