Bayesian Estimators for Bioprocess Monitoring Under Uncertainty

Abstract

The simultaneous growing interest in the application of digital solutions and the use of biological organisms such as bacteria and yeasts for the production of chemicals, has increased the necessity to apply digital solutions to biological processes, with the intent of improving production while maintaining quality standards.

The class of digital tools that can help to monitor these processes in real-time is certainly of interest. Such digital tools may carefully merge all the available information to improve the monitoring. Especially, when the qualities and quantities required are either difficult or impossible to measure directly.

The focus of this thesis is therefore the investigation and application in an experimental set-up of tools, called Soft Sensors. This was done to merge available measurements together with the physical, chemical and biological knowledge described trough mathematical relations, for monitoring the glucose consumption, which was used by the bacteria to grow.