Show simple item record

dc.contributor.authorZurita Millán, Daniel
dc.contributor.authorDelgado Prieto, Miquel
dc.contributor.authorCariño Corrales, Jesús Adolfo
dc.contributor.authorOrtega Redondo, Juan Antonio
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament d'Enginyeria Electrònica
dc.date.accessioned2017-10-27T07:55:25Z
dc.date.available2017-10-27T07:55:25Z
dc.date.issued2017-09-20
dc.identifier.citationZurita, D., Delgado Prieto, M., Cariño , J.A., Ortega, J.A. Multimodal forecasting methodology applied to industrial process monitoring. "IEEE transactions on industrial informatics", 20 Setembre 2017, p. 1-10.
dc.identifier.issn1551-3203
dc.identifier.urihttp://hdl.handle.net/2117/109279
dc.description.abstractIEEE Industrial process modelling represents a key factor to allow the future generation of industrial manufacturing plants. In this regard, accurate models of critical signals need to be designed in order to forecast process deviations. In this work a novel multimodal forecasting methodology based on adaptive dynamics packaging and codification of the process operation is proposed. First, a target signal is decomposed by means of the Empirical Mode Decomposition in order to identify the characteristics intrinsic mode functions. Second, such dynamics are packaged depending on their significance and modelling complexity. Third, the operating condition of the considered process, reflected by available auxiliary signals, is codified by means of a Self-Organizing Map and presented to the modelling structure. The forecasting structure is supported by a set of ensemble ANFIS based models, each one focused on a different set of signal dynamics. The performance and effectiveness of the proposed method is validated experimentally with industrial data from a copper rod manufacturing plant and performance comparison with classical approaches. The proposed method improves performance and generalization versus classical single model approaches.
dc.format.extent10 p.
dc.language.isoeng
dc.subjectÀrees temàtiques de la UPC::Informàtica::Aplicacions de la informàtica
dc.subject.lcshForecasting
dc.subject.lcshPackaging
dc.subject.lcshPredictive control
dc.subject.otherAdaptation models
dc.subject.otherBiological system modeling
dc.subject.otherComputational modeling
dc.subject.otherForecasting
dc.subject.otherForecasting
dc.subject.otherFuzzy neural networks
dc.subject.otherIndustrial plants
dc.subject.otherPackaging
dc.subject.otherPredictive models
dc.subject.otherPredictive models
dc.subject.otherTime series analysis
dc.subject.otherTime series analysis
dc.titleMultimodal forecasting methodology applied to industrial process monitoring
dc.typeArticle
dc.subject.lemacPrevisió
dc.subject.lemacEnvasament
dc.subject.lemacControl predictiu
dc.contributor.groupUniversitat Politècnica de Catalunya. MCIA - Motion Control and Industrial Applications Research Group
dc.identifier.doi10.1109/TII.2017.2755099
dc.description.peerreviewedPeer Reviewed
dc.relation.publisherversionhttp://ieeexplore.ieee.org/document/8048029/
dc.rights.accessOpen Access
drac.iddocument21577011
dc.description.versionPostprint (author's final draft)
upcommons.citation.authorZurita, D.; Delgado Prieto, M.; Cariño, J.A.; Ortega, J.A.
upcommons.citation.publishedtrue
upcommons.citation.publicationNameIEEE transactions on industrial informatics
upcommons.citation.startingPage1
upcommons.citation.endingPage10


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

All rights reserved. This work is protected by the corresponding intellectual and industrial property rights. Without prejudice to any existing legal exemptions, reproduction, distribution, public communication or transformation of this work are prohibited without permission of the copyright holder