Publication:
Application of Machine Learning Technique Using Support Vector Machine in Wind Turbine Fault Diagnosis

Simple item page
dc.contributor.authorBinti Shahrulhisham N.N.H.en_US
dc.contributor.authorChong K.H.en_US
dc.contributor.authorYaw C.T.en_US
dc.contributor.authorKoh S.P.en_US
dc.contributor.authorid57884510200en_US
dc.contributor.authorid36994481200en_US
dc.contributor.authorid36560884300en_US
dc.contributor.authorid57883863700en_US
dc.date.accessioned2023-05-29T09:39:54Z
dc.date.available2023-05-29T09:39:54Z
dc.date.issued2022
dc.descriptionFailure analysis; Fault detection; Gaussian distribution; Global warming; Learning algorithms; Support vector machines; Wind power; Cost-efficient; Faults diagnosis; Gaussians; Harmful gas; Machine learning techniques; Performance; Renewable energies; Support vectors machine; Wind turbine faults; Wind turbine systems; Wind turbinesen_US
dc.description.abstractWind energies are one of the most used resources worldwide and favours the economy by not emitting harmful gases that could lead to global warming. It is a cost-efficient method and environmentally friendly. Hence, explains the popularity of wind energy production over the years. Unfortunately, a minor fault could be contagious by affecting the nearby components, then a more complicated problem might arise, which may be costly. Thus, this article conducted a machine learning technique, support vector machine (SVM) to monitor the health of the wind turbine system by classifying the class of healthy data and faulty data. Some SVM types were experimented with, including Linear, Quadratic, Cubic, Fine Gaussian, Medium Gaussian, and Coarse Gaussian. Then these models were trained under different validation schemes that are cross-validation, holdout validation, and re-substitution validation as an approach to evaluate the performance of each model. In the end, Cubic SVM is proven to outperformed other models under the provision of 10-fold cross-validation with an accuracy of 98.25%. The result showed that Cubic SVM has the best performance while Linear SVM has the least accuracy among other models. Hence choosing the default value is preferred as the final product to diagnose the fault in wind turbine systems. � Published under licence by IOP Publishing Ltd.en_US
dc.description.natureFinalen_US
dc.identifier.ArtNo12017
dc.identifier.doi10.1088/1742-6596/2319/1/012017
dc.identifier.issue1
dc.identifier.scopus2-s2.0-85137684452
dc.identifier.urihttps://www.scopus.com/inward/record.uri?eid=2-s2.0-85137684452&doi=10.1088%2f1742-6596%2f2319%2f1%2f012017&partnerID=40&md5=10e925bb05953de1e873a5ef48a0f01e
dc.identifier.urihttps://irepository.uniten.edu.my/handle/123456789/27126
dc.identifier.volume2319
dc.publisherInstitute of Physicsen_US
dc.relation.ispartofAll Open Access, Gold
dc.sourceScopus
dc.sourcetitleJournal of Physics: Conference Series
dc.titleApplication of Machine Learning Technique Using Support Vector Machine in Wind Turbine Fault Diagnosisen_US
dc.typeConference Paperen_US
dspace.entity.typePublication
Files
Collections
SCOPUS