Feature Space Transformation for Fault Diagnosis of Rotating Machinery under Different Working Conditions
In recent years, various deep learning models have been developed for the fault diagnosis of rotating machines. However, in practical applications related to fault diagnosis, it is difficult to immediately implement a trained model because the distribution of source data and target domain data have...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/4/1417 |
| _version_ | 1827588927792349184 |
|---|---|
| author | Gye-Bong Jang Sung-Bae Cho |
| author_facet | Gye-Bong Jang Sung-Bae Cho |
| author_sort | Gye-Bong Jang |
| collection | DOAJ |
| description | In recent years, various deep learning models have been developed for the fault diagnosis of rotating machines. However, in practical applications related to fault diagnosis, it is difficult to immediately implement a trained model because the distribution of source data and target domain data have different distributions. Additionally, collecting failure data for various operating conditions is time consuming and expensive. In this paper, we introduce a new transformation method for the latent space between domains using the source domain and normal data of the target domain that can be easily collected. Inspired by semantic transformations in an embedded space in the field of word embedding, discrepancies between the distribution of the source and target domains are minimized by transforming the latent representation space in which fault attributes are preserved. To match the feature area and distribution, spatial attention is applied to learn the latent feature spaces, and the 1D CNN LSTM architecture is implemented to maximize the intra-class classification. The proposed model was validated for two types of rotating machines such as a dataset of rolling bearings as CWRU and a gearbox dataset of heavy machinery. Experimental results show the proposed method has higher cross-domain diagnostic accuracy than others, therefore showing reliable generalization performance in rotating machines operating under various conditions. |
| first_indexed | 2024-03-09T00:46:31Z |
| format | Article |
| id | doaj.art-37fe483e505b4a529d9e2aeb33c5875a |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T00:46:31Z |
| publishDate | 2021-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-37fe483e505b4a529d9e2aeb33c5875a2023-12-11T17:28:42ZengMDPI AGSensors1424-82202021-02-01214141710.3390/s21041417Feature Space Transformation for Fault Diagnosis of Rotating Machinery under Different Working ConditionsGye-Bong Jang0Sung-Bae Cho1Department of Computer Science, Yonsei University, Seoul 03722, KoreaDepartment of Computer Science, Yonsei University, Seoul 03722, KoreaIn recent years, various deep learning models have been developed for the fault diagnosis of rotating machines. However, in practical applications related to fault diagnosis, it is difficult to immediately implement a trained model because the distribution of source data and target domain data have different distributions. Additionally, collecting failure data for various operating conditions is time consuming and expensive. In this paper, we introduce a new transformation method for the latent space between domains using the source domain and normal data of the target domain that can be easily collected. Inspired by semantic transformations in an embedded space in the field of word embedding, discrepancies between the distribution of the source and target domains are minimized by transforming the latent representation space in which fault attributes are preserved. To match the feature area and distribution, spatial attention is applied to learn the latent feature spaces, and the 1D CNN LSTM architecture is implemented to maximize the intra-class classification. The proposed model was validated for two types of rotating machines such as a dataset of rolling bearings as CWRU and a gearbox dataset of heavy machinery. Experimental results show the proposed method has higher cross-domain diagnostic accuracy than others, therefore showing reliable generalization performance in rotating machines operating under various conditions.https://www.mdpi.com/1424-8220/21/4/1417fault diagnosisdomain adaptationattention mechanismfeature space transformationgearboxvibration measurement |
| spellingShingle | Gye-Bong Jang Sung-Bae Cho Feature Space Transformation for Fault Diagnosis of Rotating Machinery under Different Working Conditions Sensors fault diagnosis domain adaptation attention mechanism feature space transformation gearbox vibration measurement |
| title | Feature Space Transformation for Fault Diagnosis of Rotating Machinery under Different Working Conditions |
| title_full | Feature Space Transformation for Fault Diagnosis of Rotating Machinery under Different Working Conditions |
| title_fullStr | Feature Space Transformation for Fault Diagnosis of Rotating Machinery under Different Working Conditions |
| title_full_unstemmed | Feature Space Transformation for Fault Diagnosis of Rotating Machinery under Different Working Conditions |
| title_short | Feature Space Transformation for Fault Diagnosis of Rotating Machinery under Different Working Conditions |
| title_sort | feature space transformation for fault diagnosis of rotating machinery under different working conditions |
| topic | fault diagnosis domain adaptation attention mechanism feature space transformation gearbox vibration measurement |
| url | https://www.mdpi.com/1424-8220/21/4/1417 |
| work_keys_str_mv | AT gyebongjang featurespacetransformationforfaultdiagnosisofrotatingmachineryunderdifferentworkingconditions AT sungbaecho featurespacetransformationforfaultdiagnosisofrotatingmachineryunderdifferentworkingconditions |