CORTEX H2020
Published October 3, 2021 | Version v1
Presentation Open

Deep Learning-based Anomaly Detection in Nuclear Reactor Cores

Creators

  • 1. Institute of Communication and Computer Systems National Technical University of Athens
  • 2. Paul Scherrer Institute

Description

•The introduction of a deep learning methodology for the classification of different perturbation types and their position in the reactor core, using convolutional neural networks
•The performance of a complementary robustness analysis to assess the system's performance on noisy or missing data
•The assessment of the system's functionality on plant measurements obtained from the Gösgennuclear power plan in Switzerland

Files

2021_Alexandridis_MC2021_presentation1_V1.pdf

Files (656.5 kB)

Name Size Download all
md5:fe13a3eea1058186541a4408a4b038b2
656.5 kB Preview Download

Additional details

Funding

CORTEX – Core monitoring techniques and experimental validation and demonstration 754316
European Commission