A Multiscale Feature Learning Scheme Based on Deep Learning for Industrial Process Monitoring and Fault Diagnosis
Considering that industrial data exhibit nonlinearity, high dimensionality and inherent multiscale characteristics, this paper proposes an intelligent industrial process monitoring and fault diagnosis method based on the discrete wavelet transform and deep learning. First, the discrete wavelet trans...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2019-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/8871174/ |
| id |
doaj-3c1e0b1a127e4631892ab9fd3dd94763 |
|---|---|
| record_format |
Article |
| spelling |
doaj-3c1e0b1a127e4631892ab9fd3dd947632021-03-29T23:03:49ZengIEEEIEEE Access2169-35362019-01-01715118915120210.1109/ACCESS.2019.29477148871174A Multiscale Feature Learning Scheme Based on Deep Learning for Industrial Process Monitoring and Fault DiagnosisJing Yuan0https://orcid.org/0000-0001-6162-0025Ying Tian1https://orcid.org/0000-0003-0835-9731School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai, ChinaSchool of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai, ChinaConsidering that industrial data exhibit nonlinearity, high dimensionality and inherent multiscale characteristics, this paper proposes an intelligent industrial process monitoring and fault diagnosis method based on the discrete wavelet transform and deep learning. First, the discrete wavelet transform is used to present the multiscale representation of the raw data. Second, a multiscale convolutional neural network is used to extract the features at each scale, and then the extracted multiscale features are fused by the long short-term memory network to further reduce useless information and retain useful information. Finally, softmax classification is performed. The proposed method has two advantages: 1) the hierarchical learning structure with multiple pairs of convolutional and pooling layers can effectively learn nonlinear, high-dimensional fault features; and 2) the multiscale feature learning scheme can capture complementary diagnosis information at different scales. Detailed comparative studies between the proposed method and conventional methods have been carried out through the Tennessee Eastman benchmark process and the p-xylene oxidation reaction process.https://ieeexplore.ieee.org/document/8871174/Intelligent fault diagnosisindustrial processdeep learningconvolutional neural networknonlinear process monitoringindustrial application |
| collection |
DOAJ |
| language |
English |
| format |
Article |
| sources |
DOAJ |
| author |
Jing Yuan Ying Tian |
| spellingShingle |
Jing Yuan Ying Tian A Multiscale Feature Learning Scheme Based on Deep Learning for Industrial Process Monitoring and Fault Diagnosis IEEE Access Intelligent fault diagnosis industrial process deep learning convolutional neural network nonlinear process monitoring industrial application |
| author_facet |
Jing Yuan Ying Tian |
| author_sort |
Jing Yuan |
| title |
A Multiscale Feature Learning Scheme Based on Deep Learning for Industrial Process Monitoring and Fault Diagnosis |
| title_short |
A Multiscale Feature Learning Scheme Based on Deep Learning for Industrial Process Monitoring and Fault Diagnosis |
| title_full |
A Multiscale Feature Learning Scheme Based on Deep Learning for Industrial Process Monitoring and Fault Diagnosis |
| title_fullStr |
A Multiscale Feature Learning Scheme Based on Deep Learning for Industrial Process Monitoring and Fault Diagnosis |
| title_full_unstemmed |
A Multiscale Feature Learning Scheme Based on Deep Learning for Industrial Process Monitoring and Fault Diagnosis |
| title_sort |
multiscale feature learning scheme based on deep learning for industrial process monitoring and fault diagnosis |
| publisher |
IEEE |
| series |
IEEE Access |
| issn |
2169-3536 |
| publishDate |
2019-01-01 |
| description |
Considering that industrial data exhibit nonlinearity, high dimensionality and inherent multiscale characteristics, this paper proposes an intelligent industrial process monitoring and fault diagnosis method based on the discrete wavelet transform and deep learning. First, the discrete wavelet transform is used to present the multiscale representation of the raw data. Second, a multiscale convolutional neural network is used to extract the features at each scale, and then the extracted multiscale features are fused by the long short-term memory network to further reduce useless information and retain useful information. Finally, softmax classification is performed. The proposed method has two advantages: 1) the hierarchical learning structure with multiple pairs of convolutional and pooling layers can effectively learn nonlinear, high-dimensional fault features; and 2) the multiscale feature learning scheme can capture complementary diagnosis information at different scales. Detailed comparative studies between the proposed method and conventional methods have been carried out through the Tennessee Eastman benchmark process and the p-xylene oxidation reaction process. |
| topic |
Intelligent fault diagnosis industrial process deep learning convolutional neural network nonlinear process monitoring industrial application |
| url |
https://ieeexplore.ieee.org/document/8871174/ |
| work_keys_str_mv |
AT jingyuan amultiscalefeaturelearningschemebasedondeeplearningforindustrialprocessmonitoringandfaultdiagnosis AT yingtian amultiscalefeaturelearningschemebasedondeeplearningforindustrialprocessmonitoringandfaultdiagnosis AT jingyuan multiscalefeaturelearningschemebasedondeeplearningforindustrialprocessmonitoringandfaultdiagnosis AT yingtian multiscalefeaturelearningschemebasedondeeplearningforindustrialprocessmonitoringandfaultdiagnosis |
| _version_ |
1724190080400097280 |