Show simple item record

dc.contributor.authorPontes, Fabrcio Jose [UNESP]
dc.contributor.authorSilva, Messias Borges [UNESP]
dc.contributor.authorFerreira, Joao Roberto
dc.contributor.authorde Paiva, Anderson Paulo
dc.contributor.authorBalestrassi, Pedro Paulo
dc.contributor.authorSchoenhorst, Gustavo Bonnard
dc.date.accessioned2014-05-20T13:28:29Z
dc.date.available2014-05-20T13:28:29Z
dc.date.issued2010-12-01
dc.identifierhttp://dx.doi.org/10.1590/S1678-58782010000500010
dc.identifier.citationJournal of The Brazilian Society of Mechanical Sciences and Engineering. Rio de Janeiro Rj: Abcm Brazilian Soc Mechanical Sciences & Engineering, v. 32, n. 5, p. 503-510, 2010.
dc.identifier.issn1678-5878
dc.identifier.urihttp://hdl.handle.net/11449/9480
dc.description.abstractThe use of artificial neural networks for prediction in hard turning has received considerable attention in literature. An often quoted drawback of ANNs is the lack of a systematic way for the design of high performance networks. This study presents a DOE based approach for the design of ANNs of Radial Basis Function (RBF) architecture applied to surface roughness prediction in turning of AISI 52100 hardened steel. Experimental factors are the number of radial units on the hidden layer, the algorithm employed to calculate the spread factor of radial units and the algorithm employed to calculate radial function centers. DOE is employed to select levels of factors that benefit network prediction skills. Experiments with data sets of distinct sizes were conducted and network configurations leading to high performance were identified. ANN models obtained proved capable to predict roughness in accurate, precise and affordable way. Results pointed significant factors for network design and revealed that interaction effects between design parameters have significant influence on network performance for the task proposed. The work concludes that the DOE methodology constitutes a better approach to the design of RBF networks for roughness prediction than the most common trial and error approach.en
dc.description.sponsorshipFundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG)
dc.description.sponsorshipCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
dc.description.sponsorshipConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
dc.format.extent503-510
dc.language.isoeng
dc.publisherAbcm Brazilian Soc Mechanical Sciences & Engineering
dc.relation.ispartofJournal of the Brazilian Society of Mechanical Sciences and Engineering
dc.sourceWeb of Science
dc.subjectsurface roughnessen
dc.subjectdesign of experimentsen
dc.subjectradial basis function neural networksen
dc.subjecthard turningen
dc.subjectAISI 52100 hardened steelen
dc.titleA DOE based approach for the design of RBF artificial neural networks applied to prediction of surface roughness in AISI 52100 hardened Steel turningen
dc.typeArtigo
dcterms.licensehttp://www.scielo.br/revistas/jbsmse/paboutj.htm
dcterms.rightsHolderAbcm Brazilian Soc Mechanical Sciences & Engineering
dc.contributor.institutionUniversidade Estadual Paulista (UNESP)
dc.contributor.institutionUniversidade Federal de Itajubá (UNIFEI)
dc.description.affiliationUNESP Univ Estadual Paulista, Fac Engn Guaratingueta, Dept Mecan, BR-12516410 São Paulo, Brazil
dc.description.affiliationUNIFEI Universidade Federal de Itajubá (UNIFEI), Inst Eng Prod & Gestao, BR-37500903 Itajuba, MG, Brazil
dc.description.affiliationUnespUNESP Univ Estadual Paulista, Fac Engn Guaratingueta, Dept Mecan, BR-12516410 São Paulo, Brazil
dc.identifier.scieloS1678-58782010000500010
dc.identifier.wosWOS:000288384200010
dc.rights.accessRightsAcesso aberto
dc.description.sponsorshipIdFAPEMIG: PE 024/2008
unesp.campusUniversidade Estadual Paulista (UNESP), Faculdade de Engenharia, Guaratinguetápt
dc.identifier.fileS1678-58782010000500010-en.pdf
dc.identifier.lattes9507655803234261
unesp.author.lattes9507655803234261
dc.relation.ispartofjcr1.627
dc.relation.ispartofsjr0,362
Localize o texto completo

Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record