The Computing Division Technical Meetings are a platform for

  • presenting Pinboard papers under review for journal or conference publication,
  • inviting speakers who are current or prospective UKAEA collaborators at external organisations,
  • presenting work done in PhD projects funded by or co-supervised by UKAEA,
  • presenting work done during summer placements or other secondments to UKAEA.

 

If you would like to invite a speaker on a topic that would be of interest to one or more Units within the Computing Division, but is not currently collaborating on a UKAEA project, please consider nominating them for a Computing Division Cross-Disciplinary Seminar.

These meetings are normally recorded. Recordings of past meetings can be found here:

CD Technical Meetings Archive on UKAEA Sharepoint

CD Technical Meeting (ML13): Surrogate models for tokamak MHD: parameter and state inference using Ensemble Kalman Filtering

→ Europe/London
Alasdair Roy (University of Leeds)
Description
Machine Learning, Uncertainty Quantification and Data Science 

 

This talk explores the application of ensemble Kalman filtering (EnKF) to the study of tokamak instabilities, with a focus on combining data-driven approaches with reduced-order, surrogate models. While surrogates capture qualitative features of tokamak instabilities, relating them directly to experimental observations requires parameter and state inference via noise-robust data assimilation framework.

The talk begins with a brief introduction to the theory of ensemble Kalman filters and their role in data assimilation. This is followed by a description of a Python-based implementation adapted during my PhD, demonstrating how EnKF can be coupled with surrogate models for parameter and state inference. A worked example is presented to illustrate the structure and functionality of the code, and how it can be readily applied to experimental data.

Finally, the strengths and limitations of this approach are discussed when using EnKF for parameter inference.