Realising reduced order models for heat transfer problems via Data Assimilation and Machine Learning
- Academic lead
- Steven Tobias (Mathematics)
- Industrial lead
- Wayne Arter, Culham Centre for Fusion Energy
- Co-supervisor(s)
- Phil Livermore (Earth & Environment), Chris Jones (Mathematics)
- Project themes
- Energy and Transport, Underpinning Methods for Fluid Dynamics
This project is to use data-driven methods and machine learning to derive reduced models for fluid flows. In particular, we shall look at models for heat transport. Detailed fluid and plasma simulation is often so computationally expensive that it restricts the capability to design and control experiments. We shall use data-driven models and machine learning techniques to derive models that could rapidly indicate quantitatively which new experiments ought to be conducted, and help predict their behaviour. The work is of general applicability in many fields of fluid modelling and will be a step towards the ITER goal of being able to predict and control the production of nuclear fusion energy from the tokamak device to within a known degree of precision