Improved Ensemble Learning for Wind Turbine Main Bearing Fault Diagnosis
| Autor: | Olga Porro, Jose Sepulveda, Mattia Beretta, Jordi Cusidó, Yolanda Vidal |
|---|---|
| Přispěvatelé: | Universitat Politècnica de Catalunya. Doctorat en Enginyeria Ambiental, Universitat Politècnica de Catalunya. Departament de Matemàtiques, Universitat Politècnica de Catalunya. Departament d'Enginyeria de Projectes i de la Construcció, Universitat Politècnica de Catalunya. CoDAlab - Control, Modelització, Identificació i Aplicacions |
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
Technology
main bearing Computer science QH301-705.5 Real SCADA data QC1-999 Matemàtiques i estadística::Matemàtica aplicada a les ciències [Àrees temàtiques de la UPC] real SCADA data ComputerApplications_COMPUTERSINOTHERSYSTEMS Fault (power engineering) Turbine Aerogeneradors SCADA Wind turbines General Materials Science Biology (General) Instrumentation QD1-999 Fluid Flow and Transfer Processes Wind power Artificial neural network fault prognosis WT business.industry Process Chemistry and Technology Physics General Engineering Main bearing Engineering (General). Civil engineering (General) Ensemble learning Computer Science Applications Reliability engineering SCADA (Programes d'ordinador) Fault prognosis Chemistry Anomaly detection Energies::Energia eòlica [Àrees temàtiques de la UPC] normality model TA1-2040 business |
| Zdroj: | Applied Sciences, Vol 11, Iss 7523, p 7523 (2021) Applied Sciences Volume 11 Issue 16 UPCommons. Portal del coneixement obert de la UPC Universitat Politècnica de Catalunya (UPC) |
| ISSN: | 2076-3417 |
| Popis: | The goal of this paper is to develop, implement, and validate a methodology for wind turbines’ main bearing fault prediction based on an ensemble of an artificial neural network (normality model designed at turbine level) and an isolation forest (anomaly detection model designed at wind park level) algorithms trained only on SCADA data. The normal behavior and the anomalous samples of the wind turbines are identified and several interpretable indicators are proposed based on the predictions of these algorithms, to provide the wind park operators with understandable information with enough time to plan operations ahead and avoid unexpected costs. The stated methodology is validated in a real underproduction wind park composed by 18 wind turbines. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|