Wind turbine fault diagnosis based on transfer learning and convolutional autoencoder with small-scale data

Condition monitoring and fault diagnosis for wind turbines can effectively reduce the impact of failures. However, many wind turbines cannot establish fault diagnosis models due to insufficient data. The operational data of similar wind turbines usually contain some universal information about failu...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	Renewable energy Ročník 171; s. 103 - 115
Hlavní autoři:	Li, Yanting, Jiang, Wenbo, Zhang, Guangyao, Shu, Lianjie
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	Elsevier Ltd 01.06.2021
Témata:	Convolutional autoencoder exhibitions Fault diagnosis learning monitoring renewable energy sources Small-scale data Transfer learning Wind turbine wind turbines ANN CNN LSTM AE Transfer learning CAE SSI GP SVM Convolutional autoencoder TTL Fault diagnosis DAN DT TCA RF WWEA AE-TL CAE-TL Small-scale data Wind turbine DaNN SCL SCADA
ISSN:	0960-1481, 1879-0682
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Popis
Shrnutí:	Condition monitoring and fault diagnosis for wind turbines can effectively reduce the impact of failures. However, many wind turbines cannot establish fault diagnosis models due to insufficient data. The operational data of similar wind turbines usually contain some universal information about failure properties. In order to make full use of these useful information, a fault diagnosis method based on parameter-based transfer learning and convolutional autoencoder (CAE) for wind turbines with small-scale data is proposed in this paper. The proposed method can transfer knowledge from similar wind turbines to the target wind turbine. The performance of the proposed method is analyzed and compared to other transfer/non-transfer methods. The comparison results show that the proposed method has advantages in diagnosing faults for wind turbines with small-scale data. •Suggest a new model based on transfer learning for wind turbine diagnosis with small-scale data.•The suggested model can take the operational information from other wind turbines into account.•The comparison results favor the new model.
Bibliografie:	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23
ISSN:	0960-1481 1879-0682
DOI:	10.1016/j.renene.2021.01.143