IMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPU
Autor(a) principal: | |
---|---|
Data de Publicação: | 2017 |
Outros Autores: | , , |
Tipo de documento: | Artigo de conferência |
Idioma: | eng |
Título da fonte: | Repositório Institucional do IEN |
Texto Completo: | http://carpedien.ien.gov.br:8080/handle/ien/2152 |
Resumo: | Recently, the use of mobile devices has been proposed for dose assessment during nuclear accidents. The idea is to support field teams, providing an approximated estimation of the dose distribution map in the vicinity of the nuclear power plant (NPP), without needing to be connected to the NPP systems. In order to provide such stand-alone execution, the use of artificial neural networks (ANN) has been proposed in substitution of the complex and time consuming physical models executed by the atmospheric dispersion radionuclide (ADR) system. One limitation observed on such approach is the very time-consuming training of the ANNs. Moreover, if the number of input parameters increases the performance of standard ANNs, like Multilayer-Perceptron (MLP) with backpropagation training, is affected leading to unreasonable training time. To improve learning, allowing better dose estimations, more complex ANN architectures are required. ANNs with many layers (much more than a typical number of layers), referred to as Deep Neural Networks (DNN), for example, have demonstrating to achieve better results. On the other hand, the training of such ANNs is very much slow. In order to allow the use of such DNNs in a reasonable training time, a parallel programming solution, using Graphic Processing Units (GPU) and Computing Unified Device Architecture (CUDA) is proposed. This work focuses on the study of computational technologies for improvement of the ANNs to be used in the mobile application, as well as their training algorithms. |
id |
IEN_0e88b7f74d95712ae6e6f18db44e726e |
---|---|
oai_identifier_str |
oai:carpedien.ien.gov.br:ien/2152 |
network_acronym_str |
IEN |
network_name_str |
Repositório Institucional do IEN |
spelling |
DESTERRO, Filipe S. M.ALMEIDA, Adino A. H.PEREIRA, Claudio M. N. A.Instituto de Engenharia Nuclear (IEN)Instituto de Engenharia Nuclear (IEN)Instituto de Engenharia Nuclear (IEN)2018-02-08T13:42:45Z2018-02-08T13:42:45Z2017-10http://carpedien.ien.gov.br:8080/handle/ien/2152Submitted by Vanessa Silva (vanessacapucho.uerj@gmail.com) on 2018-02-08T13:42:44Z No. of bitstreams: 1 ARTIGO INAC 8.pdf: 234185 bytes, checksum: 4a95fa6cc92b9fd7675b44acb9c695cf (MD5)Made available in DSpace on 2018-02-08T13:42:45Z (GMT). No. of bitstreams: 1 ARTIGO INAC 8.pdf: 234185 bytes, checksum: 4a95fa6cc92b9fd7675b44acb9c695cf (MD5) Previous issue date: 2017-10Recently, the use of mobile devices has been proposed for dose assessment during nuclear accidents. The idea is to support field teams, providing an approximated estimation of the dose distribution map in the vicinity of the nuclear power plant (NPP), without needing to be connected to the NPP systems. In order to provide such stand-alone execution, the use of artificial neural networks (ANN) has been proposed in substitution of the complex and time consuming physical models executed by the atmospheric dispersion radionuclide (ADR) system. One limitation observed on such approach is the very time-consuming training of the ANNs. Moreover, if the number of input parameters increases the performance of standard ANNs, like Multilayer-Perceptron (MLP) with backpropagation training, is affected leading to unreasonable training time. To improve learning, allowing better dose estimations, more complex ANN architectures are required. ANNs with many layers (much more than a typical number of layers), referred to as Deep Neural Networks (DNN), for example, have demonstrating to achieve better results. On the other hand, the training of such ANNs is very much slow. In order to allow the use of such DNNs in a reasonable training time, a parallel programming solution, using Graphic Processing Units (GPU) and Computing Unified Device Architecture (CUDA) is proposed. This work focuses on the study of computational technologies for improvement of the ANNs to be used in the mobile application, as well as their training algorithms.engInstituto de Engenharia NuclearIENBrasilnuclear accidentsIMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPUinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/conferenceObject2017info:eu-repo/semantics/openAccessreponame:Repositório Institucional do IENinstname:Instituto de Engenharia Nuclearinstacron:IENLICENSElicense.txtlicense.txttext/plain; charset=utf-81748http://carpedien.ien.gov.br:8080/xmlui/bitstream/ien/2152/2/license.txt8a4605be74aa9ea9d79846c1fba20a33MD52ORIGINALARTIGO INAC 8.pdfARTIGO INAC 8.pdfapplication/pdf234185http://carpedien.ien.gov.br:8080/xmlui/bitstream/ien/2152/1/ARTIGO+INAC+8.pdf4a95fa6cc92b9fd7675b44acb9c695cfMD51ien/2152oai:carpedien.ien.gov.br:ien/21522018-02-08 11:42:45.017Dspace IENlsales@ien.gov.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 |
dc.title.pt_BR.fl_str_mv |
IMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPU |
title |
IMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPU |
spellingShingle |
IMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPU DESTERRO, Filipe S. M. nuclear accidents |
title_short |
IMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPU |
title_full |
IMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPU |
title_fullStr |
IMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPU |
title_full_unstemmed |
IMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPU |
title_sort |
IMPROVEMENT OF RADIATION DOSE ESTIMATION DUE TO NUCLEAR ACCIDENTS USING DEEP NEURAL NETWORK AND GPU |
author |
DESTERRO, Filipe S. M. |
author_facet |
DESTERRO, Filipe S. M. ALMEIDA, Adino A. H. PEREIRA, Claudio M. N. A. Instituto de Engenharia Nuclear (IEN) |
author_role |
author |
author2 |
ALMEIDA, Adino A. H. PEREIRA, Claudio M. N. A. Instituto de Engenharia Nuclear (IEN) |
author2_role |
author author author |
dc.contributor.author.fl_str_mv |
DESTERRO, Filipe S. M. ALMEIDA, Adino A. H. PEREIRA, Claudio M. N. A. Instituto de Engenharia Nuclear (IEN) Instituto de Engenharia Nuclear (IEN) Instituto de Engenharia Nuclear (IEN) |
dc.subject.por.fl_str_mv |
nuclear accidents |
topic |
nuclear accidents |
dc.description.abstract.por.fl_txt_mv |
Recently, the use of mobile devices has been proposed for dose assessment during nuclear accidents. The idea is to support field teams, providing an approximated estimation of the dose distribution map in the vicinity of the nuclear power plant (NPP), without needing to be connected to the NPP systems. In order to provide such stand-alone execution, the use of artificial neural networks (ANN) has been proposed in substitution of the complex and time consuming physical models executed by the atmospheric dispersion radionuclide (ADR) system. One limitation observed on such approach is the very time-consuming training of the ANNs. Moreover, if the number of input parameters increases the performance of standard ANNs, like Multilayer-Perceptron (MLP) with backpropagation training, is affected leading to unreasonable training time. To improve learning, allowing better dose estimations, more complex ANN architectures are required. ANNs with many layers (much more than a typical number of layers), referred to as Deep Neural Networks (DNN), for example, have demonstrating to achieve better results. On the other hand, the training of such ANNs is very much slow. In order to allow the use of such DNNs in a reasonable training time, a parallel programming solution, using Graphic Processing Units (GPU) and Computing Unified Device Architecture (CUDA) is proposed. This work focuses on the study of computational technologies for improvement of the ANNs to be used in the mobile application, as well as their training algorithms. |
description |
Recently, the use of mobile devices has been proposed for dose assessment during nuclear accidents. The idea is to support field teams, providing an approximated estimation of the dose distribution map in the vicinity of the nuclear power plant (NPP), without needing to be connected to the NPP systems. In order to provide such stand-alone execution, the use of artificial neural networks (ANN) has been proposed in substitution of the complex and time consuming physical models executed by the atmospheric dispersion radionuclide (ADR) system. One limitation observed on such approach is the very time-consuming training of the ANNs. Moreover, if the number of input parameters increases the performance of standard ANNs, like Multilayer-Perceptron (MLP) with backpropagation training, is affected leading to unreasonable training time. To improve learning, allowing better dose estimations, more complex ANN architectures are required. ANNs with many layers (much more than a typical number of layers), referred to as Deep Neural Networks (DNN), for example, have demonstrating to achieve better results. On the other hand, the training of such ANNs is very much slow. In order to allow the use of such DNNs in a reasonable training time, a parallel programming solution, using Graphic Processing Units (GPU) and Computing Unified Device Architecture (CUDA) is proposed. This work focuses on the study of computational technologies for improvement of the ANNs to be used in the mobile application, as well as their training algorithms. |
publishDate |
2017 |
dc.date.issued.fl_str_mv |
2017-10 |
dc.date.accessioned.fl_str_mv |
2018-02-08T13:42:45Z |
dc.date.available.fl_str_mv |
2018-02-08T13:42:45Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/conferenceObject |
status_str |
publishedVersion |
format |
conferenceObject |
dc.identifier.uri.fl_str_mv |
http://carpedien.ien.gov.br:8080/handle/ien/2152 |
url |
http://carpedien.ien.gov.br:8080/handle/ien/2152 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.publisher.none.fl_str_mv |
Instituto de Engenharia Nuclear |
dc.publisher.initials.fl_str_mv |
IEN |
dc.publisher.country.fl_str_mv |
Brasil |
publisher.none.fl_str_mv |
Instituto de Engenharia Nuclear |
dc.source.none.fl_str_mv |
reponame:Repositório Institucional do IEN instname:Instituto de Engenharia Nuclear instacron:IEN |
reponame_str |
Repositório Institucional do IEN |
collection |
Repositório Institucional do IEN |
instname_str |
Instituto de Engenharia Nuclear |
instacron_str |
IEN |
institution |
IEN |
bitstream.url.fl_str_mv |
http://carpedien.ien.gov.br:8080/xmlui/bitstream/ien/2152/2/license.txt http://carpedien.ien.gov.br:8080/xmlui/bitstream/ien/2152/1/ARTIGO+INAC+8.pdf |
bitstream.checksum.fl_str_mv |
8a4605be74aa9ea9d79846c1fba20a33 4a95fa6cc92b9fd7675b44acb9c695cf |
bitstream.checksumAlgorithm.fl_str_mv |
MD5 MD5 |
repository.name.fl_str_mv |
Dspace IEN |
repository.mail.fl_str_mv |
lsales@ien.gov.br |
_version_ |
1656026990445068288 |