Data assimilation for nuclear cores

The studies on data assimilation for neutronic are the result of collaboration between CERFACS
and EDF R&D started in 2003. The goal of the actual project ARTEMIS (2008-2010) is to complete and
improve the very good results that have been obtained since 2003. Moreover, the framework of the
present studies gets the specificity to use the newest nuclear core evaluation core developed by EDF
(COCAGNE)

Compared to climatic model that are dealing with millions of points, the study of nuclear core are dealing with at most few tenth thousand of points. This reduce size of the system has allowed to make some specific and original's studies

Through data assimilation, two main classes of problem are addressed:
  • Optimal parameter determination
  • Field reconstruction

The study on optimal parameter determination lead to new ideas for modelling some parts of the
core as well as studies about treatment of non linearity in data assimilation.

About the field reconstruction many studies have been done on the topic itself as well as on some
related topics. Though these work some specific method to deal with heterogeneous instruments
have been developed. Also, the sensitivity of data assimilation to instrumental set up and to various
parameters has been studied. From those results it have been developed some techniques to
optimize instruments' locations within the core.

Beyond the studies dedicated to nuclear core these studies are a source of results that can be
applied to any physical system involving few tenth thousands variables

Most of those studies as well as their extension toward nuclear core dynamic will go on in a new
project in collaboration with EDF R&D. This project of greater size (ARTEMIS II) will cover the 2011-
2015 periods.



Back
CNESEADSEDFMeteo FranceONERASAFRANTotal
English | French | Intranet | FTP | Site Map | Legal Information | © CERFACS 2009 | Conception: CERFACS - Oréalys