Conference paper Open Access

General Regression Neural Networks for the Concurrent, Timely and Reliable Identification of Detector Malfunctions and/or Nuclear Reactor Deviations from Steady-State Operation

Tatiana Tambouratzis; Dionysios Chionis; Abdelhamid Dokhane


MARC21 XML Export

<?xml version='1.0' encoding='UTF-8'?>
<record xmlns="http://www.loc.gov/MARC21/slim">
  <leader>00000nam##2200000uu#4500</leader>
  <datafield tag="041" ind1=" " ind2=" ">
    <subfield code="a">eng</subfield>
  </datafield>
  <datafield tag="653" ind1=" " ind2=" ">
    <subfield code="a">The analysis and understanding of the neutron flux (NF) signals of nuclear reactors (NRs) is imperative for ensuring safe and optimal (expressed in terms of minimal fuel use for maximal energy production) on-line NR operation. The NF perturbations are of particular interest, as they provide detailed information concerning the instantaneous changes in NR operation/status. In this piece of research, general regression artificial neural networks (GRNNs) are proposed for concurrently identifying NR deviations from steady-state operation as well as neutron detector (ND) malfunctions in a timely, reliable and efficient manner. On the one hand, the use of (a) raw, minimalistic NF signals and (b) complementary signal encodings – derived from pertinent and limited in size ND configurations – of the problem space, renders the proposed approach timely/efficient, modular as well as flexible. On the other hand, the GRNN characteristics of (i) transparency of construction, (ii) low computational (time/space) complexity of training and testing, (iii) accuracy, consistency and good generalization in the identification of the cause(s) behind deviating-from-normal NR behaviour and (iv) efficient operation and partial only GRNN retraining following modification of the training set, support the use of the proposed methodology. It is envisaged that, by appropriately combining the responses derived from different GRNNs, both accuracy and sensitivity of deviation detection as well as of malfunction localization shall be further improved at minimal additional computational load</subfield>
  </datafield>
  <controlfield tag="005">20220816073540.0</controlfield>
  <controlfield tag="001">3047075</controlfield>
  <datafield tag="711" ind1=" " ind2=" ">
    <subfield code="g">SSCI 2018</subfield>
    <subfield code="a">IEEE Symposium Series on Computational Intelligence</subfield>
    <subfield code="c">Bengaluru, India</subfield>
  </datafield>
  <datafield tag="700" ind1=" " ind2=" ">
    <subfield code="u">Paul Scherrer Institute</subfield>
    <subfield code="a">Dionysios Chionis</subfield>
  </datafield>
  <datafield tag="700" ind1=" " ind2=" ">
    <subfield code="u">Paul Scherrer Institute</subfield>
    <subfield code="a">Abdelhamid Dokhane</subfield>
  </datafield>
  <datafield tag="856" ind1="4" ind2=" ">
    <subfield code="s">525663</subfield>
    <subfield code="z">md5:9319a0d4d2ce0fa3e6fb107dfbd8cd93</subfield>
    <subfield code="u">https://zenodo.org/record/3047075/files/2018_Tambouratzis_SSCI2018_paper_V1.pdf</subfield>
  </datafield>
  <datafield tag="542" ind1=" " ind2=" ">
    <subfield code="l">open</subfield>
  </datafield>
  <datafield tag="260" ind1=" " ind2=" ">
    <subfield code="c">2018-11-18</subfield>
  </datafield>
  <datafield tag="909" ind1="C" ind2="O">
    <subfield code="p">openaire</subfield>
    <subfield code="o">oai:zenodo.org:3047075</subfield>
  </datafield>
  <datafield tag="100" ind1=" " ind2=" ">
    <subfield code="u">University of Piraeus</subfield>
    <subfield code="a">Tatiana Tambouratzis</subfield>
  </datafield>
  <datafield tag="245" ind1=" " ind2=" ">
    <subfield code="a">General Regression Neural Networks for the Concurrent, Timely and Reliable Identification of Detector Malfunctions and/or Nuclear Reactor Deviations from Steady-State Operation</subfield>
  </datafield>
  <datafield tag="536" ind1=" " ind2=" ">
    <subfield code="c">754316</subfield>
    <subfield code="a">Core monitoring techniques and experimental validation and demonstration</subfield>
  </datafield>
  <datafield tag="540" ind1=" " ind2=" ">
    <subfield code="u">https://creativecommons.org/licenses/by/4.0/legalcode</subfield>
    <subfield code="a">Creative Commons Attribution 4.0 International</subfield>
  </datafield>
  <datafield tag="650" ind1="1" ind2="7">
    <subfield code="a">cc-by</subfield>
    <subfield code="2">opendefinition.org</subfield>
  </datafield>
  <datafield tag="520" ind1=" " ind2=" ">
    <subfield code="a">&lt;p&gt;The analysis and understanding of the neutron flux (NF) signals of nuclear reactors (NRs) is imperative for ensuring safe and optimal (expressed in terms of minimal fuel use for maximal energy production) on-line NR operation. The NF perturbations are of particular interest, as they provide detailed information concerning the instantaneous changes in NR operation/status. In this piece of research, general regression artificial neural networks (GRNNs) are proposed for concurrently identifying NR deviations from steady-state operation as well as neutron detector (ND) malfunctions in a timely, reliable and efficient manner. On the one hand, the use of (a) raw, minimalistic NF signals and (b) complementary signal encodings &amp;ndash; derived from pertinent and limited in size ND configurations &amp;ndash; of the problem space, renders the proposed approach timely/efficient, modular as well as flexible. On the other hand, the GRNN characteristics of (i) transparency of construction, (ii) low computational (time/space) complexity of training and testing, (iii) accuracy, consistency and good generalization in the identification of the cause(s) behind deviating-from-normal NR behaviour and (iv) efficient operation and partial only GRNN retraining following modification of the training set, support the use of the proposed methodology. It is envisaged that, by appropriately combining the responses derived from different GRNNs, both accuracy and sensitivity of deviation detection as well as of malfunction localization shall be further improved at minimal additional computational load.&lt;/p&gt;</subfield>
  </datafield>
  <datafield tag="773" ind1=" " ind2=" ">
    <subfield code="n">doi</subfield>
    <subfield code="i">isVersionOf</subfield>
    <subfield code="a">10.5281/zenodo.3047074</subfield>
  </datafield>
  <datafield tag="024" ind1=" " ind2=" ">
    <subfield code="a">10.5281/zenodo.3047075</subfield>
    <subfield code="2">doi</subfield>
  </datafield>
  <datafield tag="980" ind1=" " ind2=" ">
    <subfield code="a">publication</subfield>
    <subfield code="b">conferencepaper</subfield>
  </datafield>
</record>
42
74
views
downloads
All versions This version
Views 4242
Downloads 7474
Data volume 38.9 MB38.9 MB
Unique views 3838
Unique downloads 7272

Share

Cite as