Bearing Small Sample Fault Diagnosis based on Data Generation and Transfer Learning
Aiming at the problem of limited fault diagnosis performance caused by a single source of historical operating data for wind turbine bearings and a small amount of data， a small sample fault diagnosis method of bearings based on data generation and transfer learning is proposed. First of all， for th...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Editorial Office of Journal of Mechanical Transmission
2020-01-01
|
| Series: | Jixie chuandong |
| Subjects: | |
| Online Access: | http://www.jxcd.net.cn/thesisDetails#10.16578/j.issn.1004.2539.2020.11.023 |
| _version_ | 1797819586001764352 |
|---|---|
| author | Dinghui Wu Qin Fang Chuyi Wu |
| author_facet | Dinghui Wu Qin Fang Chuyi Wu |
| author_sort | Dinghui Wu |
| collection | DOAJ |
| description | Aiming at the problem of limited fault diagnosis performance caused by a single source of historical operating data for wind turbine bearings and a small amount of data， a small sample fault diagnosis method of bearings based on data generation and transfer learning is proposed. First of all， for the problems of class imbalance and data scarcity in the wind turbine bearing dataset， a data generation method based on a gate mechanism is proposed. The blade-end data coaxial with the drive-end of the bearing is used as a template to generate a sufficient amount of generated data. The data is used as the source dataset. Then， according to the time series correlation of the data and the small sample application scenario， a transfer learning method based on one-dimensional convolutional neural network （1DCNN） and bidirectional gated recurrent unit （BiGRU） is proposed. First， the source dataset is trained on the training network to obtain the source model， and then a small amount of drive-end data is used as the target dataset to fine-tuning to obtain the target model. Finally， the Softmax function is used for fault diagnosis on the output of the fully-connected layer of the target model. Experiments show that the proposed fault detection method has an average accuracy of 99.67% in the scenario of small sample data of the target dataset， so the obvious classification effect and strong generalization ability could be seen. |
| first_indexed | 2024-03-13T09:24:50Z |
| format | Article |
| id | doaj.art-0e82b6c9b8fa4dc793bd4831092a2949 |
| institution | Directory Open Access Journal |
| issn | 1004-2539 |
| language | zho |
| last_indexed | 2024-03-13T09:24:50Z |
| publishDate | 2020-01-01 |
| publisher | Editorial Office of Journal of Mechanical Transmission |
| record_format | Article |
| series | Jixie chuandong |
| spelling | doaj.art-0e82b6c9b8fa4dc793bd4831092a29492023-05-26T09:35:25ZzhoEditorial Office of Journal of Mechanical TransmissionJixie chuandong1004-25392020-01-014413914429792911Bearing Small Sample Fault Diagnosis based on Data Generation and Transfer LearningDinghui WuQin FangChuyi WuAiming at the problem of limited fault diagnosis performance caused by a single source of historical operating data for wind turbine bearings and a small amount of data， a small sample fault diagnosis method of bearings based on data generation and transfer learning is proposed. First of all， for the problems of class imbalance and data scarcity in the wind turbine bearing dataset， a data generation method based on a gate mechanism is proposed. The blade-end data coaxial with the drive-end of the bearing is used as a template to generate a sufficient amount of generated data. The data is used as the source dataset. Then， according to the time series correlation of the data and the small sample application scenario， a transfer learning method based on one-dimensional convolutional neural network （1DCNN） and bidirectional gated recurrent unit （BiGRU） is proposed. First， the source dataset is trained on the training network to obtain the source model， and then a small amount of drive-end data is used as the target dataset to fine-tuning to obtain the target model. Finally， the Softmax function is used for fault diagnosis on the output of the fully-connected layer of the target model. Experiments show that the proposed fault detection method has an average accuracy of 99.67% in the scenario of small sample data of the target dataset， so the obvious classification effect and strong generalization ability could be seen.http://www.jxcd.net.cn/thesisDetails#10.16578/j.issn.1004.2539.2020.11.023Wind turbine bearing;Small sample;Data generation;Gate mechanism;Transfer learning;Fine-tuning;Fault diagnosis |
| spellingShingle | Dinghui Wu Qin Fang Chuyi Wu Bearing Small Sample Fault Diagnosis based on Data Generation and Transfer Learning Jixie chuandong Wind turbine bearing;Small sample;Data generation;Gate mechanism;Transfer learning;Fine-tuning;Fault diagnosis |
| title | Bearing Small Sample Fault Diagnosis based on Data Generation and Transfer Learning |
| title_full | Bearing Small Sample Fault Diagnosis based on Data Generation and Transfer Learning |
| title_fullStr | Bearing Small Sample Fault Diagnosis based on Data Generation and Transfer Learning |
| title_full_unstemmed | Bearing Small Sample Fault Diagnosis based on Data Generation and Transfer Learning |
| title_short | Bearing Small Sample Fault Diagnosis based on Data Generation and Transfer Learning |
| title_sort | bearing small sample fault diagnosis based on data generation and transfer learning |
| topic | Wind turbine bearing;Small sample;Data generation;Gate mechanism;Transfer learning;Fine-tuning;Fault diagnosis |
| url | http://www.jxcd.net.cn/thesisDetails#10.16578/j.issn.1004.2539.2020.11.023 |
| work_keys_str_mv | AT dinghuiwu bearingsmallsamplefaultdiagnosisbasedondatagenerationandtransferlearning AT qinfang bearingsmallsamplefaultdiagnosisbasedondatagenerationandtransferlearning AT chuyiwu bearingsmallsamplefaultdiagnosisbasedondatagenerationandtransferlearning |