Data-Efficient Product and Process Optimisation with Bayesian Hybrid Models

Project ID: 2531bd1630

(You will need this ID for your application)

Research Theme: Engineering

Research Area(s): Manufacturing technologies
Process systemscomponents and integration
Engineering design

UCL Lead department: Chemical Engineering

Department Website

Lead Supervisor: Eike Cramer

Project Summary:

One of the persistent challenges in chemical engineering research is the manufacturing of specialised, high-value products such as pharmaceuticals or functional polymers. Rigorous modelling and optimisation can help find efficient process designs and operating conditions while maintaining quality standards. However, acquiring detailed process insights for extensive modelling is often prohibitively expensive for products produced in small quantities. As a result, manufacturing processes are often run in suboptimal conditions, leading to high costs for process operation and downstream processing.

This project will utilise innovative machine learning methods and tools from process systems engineering to simultaneously optimise the product quality and the manufacturing process. The project will further focus on batch processes as the most common production process for small quantity products. Key components of this work are the formulation of mathematical optimisation problems based on mechanistic and machine learning models, developing modelling and system identification methodologies with a focus on data efficiency, as well as demonstrating the efficacy of the results on relevant case studies.

The successful candidate should have a keen interest in computational tools, numerical process optimisation, and a good understanding of chemical engineering fundamentals, including process dynamics. Through the completion of this project, the candidate will become a multi-domain expert with expertise in chemical engineering and computational methods, including machine learning, simulation, and optimisation.

You will be working with Dr Eike Cramer and Prof. Federico Galvanin at the Department of Chemical Engineering and the Seargent Centre for Process Systems Engineering.