In the past few years, reactor thermal-hydraulic (T-H) study has advanced with the support of machine learning (ML) in many aspects, including automated experimental data analysis, data-driven prediction for important reactor thermal-fluid phenomena, and surrogate modeling and uncertainty quantification for reactor system codes. ML also showed promising potential to expand reactor T-H to a wider range of applications to better support advanced reactor deployment, such as integrated multi-physics modeling and digital twin. On the other hand, ML in T-H study has its unique challenges, from data availability and quality to model transparency and interpretability. In this panel session, experts from different institutes with a diverse background will share their experience and perspectives on ML for T-H study, including recent progresses, existing challenges and potential solutions, and future opportunities.

Panelists

  • Prashant Jain (ORNL)
  • Xu Wu (NCSU)
  • Juliana Pacheco Duarte (Virginia Tech)
  • Yang Liu (ANL)
  • Pat Everett (Oklo Inc)

Resources

To access session resources, you must be logged in and registered for the meeting.

Register NowLog In

Discussion

There are 2 comments in this discussion.

To join the conversation, you must be logged in and registered for the meeting.

Register NowLog In