Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault Diagnosis
Abstract In machinery fault diagnosis, labeled data are always difficult or even impossible to obtain. Transfer learning can leverage related fault diagnosis knowledge from fully labeled source domain to enhance the fault diagnosis performance in sparsely labeled or unlabeled target domain, which ha...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2021-06-01
|
| Series: | Chinese Journal of Mechanical Engineering |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s10033-021-00566-3 |
| _version_ | 1819015896970035200 |
|---|---|
| author | Yixiao Liao Ruyi Huang Jipu Li Zhuyun Chen Weihua Li |
| author_facet | Yixiao Liao Ruyi Huang Jipu Li Zhuyun Chen Weihua Li |
| author_sort | Yixiao Liao |
| collection | DOAJ |
| description | Abstract In machinery fault diagnosis, labeled data are always difficult or even impossible to obtain. Transfer learning can leverage related fault diagnosis knowledge from fully labeled source domain to enhance the fault diagnosis performance in sparsely labeled or unlabeled target domain, which has been widely used for cross domain fault diagnosis. However, existing methods focus on either marginal distribution adaptation (MDA) or conditional distribution adaptation (CDA). In practice, marginal and conditional distributions discrepancies both have significant but different influences on the domain divergence. In this paper, a dynamic distribution adaptation based transfer network (DDATN) is proposed for cross domain bearing fault diagnosis. DDATN utilizes the proposed instance-weighted dynamic maximum mean discrepancy (IDMMD) for dynamic distribution adaptation (DDA), which can dynamically estimate the influences of marginal and conditional distribution and adapt target domain with source domain. The experimental evaluation on cross domain bearing fault diagnosis demonstrates that DDATN can outperformance the state-of-the-art cross domain fault diagnosis methods. |
| first_indexed | 2024-12-21T02:39:02Z |
| format | Article |
| id | doaj.art-21059d4d6180411282ea151f668d8075 |
| institution | Directory Open Access Journal |
| issn | 1000-9345 2192-8258 |
| language | English |
| last_indexed | 2024-12-21T02:39:02Z |
| publishDate | 2021-06-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | Chinese Journal of Mechanical Engineering |
| spelling | doaj.art-21059d4d6180411282ea151f668d80752022-12-21T19:18:43ZengSpringerOpenChinese Journal of Mechanical Engineering1000-93452192-82582021-06-0134111010.1186/s10033-021-00566-3Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault DiagnosisYixiao Liao0Ruyi Huang1Jipu Li2Zhuyun Chen3Weihua Li4School of Mechanical & Automotive Engineering, South Chine University of TechnologySchool of Mechanical & Automotive Engineering, South Chine University of TechnologySchool of Mechanical & Automotive Engineering, South Chine University of TechnologySchool of Mechanical & Automotive Engineering, South Chine University of TechnologySchool of Mechanical & Automotive Engineering, South Chine University of TechnologyAbstract In machinery fault diagnosis, labeled data are always difficult or even impossible to obtain. Transfer learning can leverage related fault diagnosis knowledge from fully labeled source domain to enhance the fault diagnosis performance in sparsely labeled or unlabeled target domain, which has been widely used for cross domain fault diagnosis. However, existing methods focus on either marginal distribution adaptation (MDA) or conditional distribution adaptation (CDA). In practice, marginal and conditional distributions discrepancies both have significant but different influences on the domain divergence. In this paper, a dynamic distribution adaptation based transfer network (DDATN) is proposed for cross domain bearing fault diagnosis. DDATN utilizes the proposed instance-weighted dynamic maximum mean discrepancy (IDMMD) for dynamic distribution adaptation (DDA), which can dynamically estimate the influences of marginal and conditional distribution and adapt target domain with source domain. The experimental evaluation on cross domain bearing fault diagnosis demonstrates that DDATN can outperformance the state-of-the-art cross domain fault diagnosis methods.https://doi.org/10.1186/s10033-021-00566-3Cross domain fault diagnosisDynamic distribution adaptationInstance-weighted dynamic MMDTransfer learning |
| spellingShingle | Yixiao Liao Ruyi Huang Jipu Li Zhuyun Chen Weihua Li Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault Diagnosis Chinese Journal of Mechanical Engineering Cross domain fault diagnosis Dynamic distribution adaptation Instance-weighted dynamic MMD Transfer learning |
| title | Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault Diagnosis |
| title_full | Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault Diagnosis |
| title_fullStr | Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault Diagnosis |
| title_full_unstemmed | Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault Diagnosis |
| title_short | Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault Diagnosis |
| title_sort | dynamic distribution adaptation based transfer network for cross domain bearing fault diagnosis |
| topic | Cross domain fault diagnosis Dynamic distribution adaptation Instance-weighted dynamic MMD Transfer learning |
| url | https://doi.org/10.1186/s10033-021-00566-3 |
| work_keys_str_mv | AT yixiaoliao dynamicdistributionadaptationbasedtransfernetworkforcrossdomainbearingfaultdiagnosis AT ruyihuang dynamicdistributionadaptationbasedtransfernetworkforcrossdomainbearingfaultdiagnosis AT jipuli dynamicdistributionadaptationbasedtransfernetworkforcrossdomainbearingfaultdiagnosis AT zhuyunchen dynamicdistributionadaptationbasedtransfernetworkforcrossdomainbearingfaultdiagnosis AT weihuali dynamicdistributionadaptationbasedtransfernetworkforcrossdomainbearingfaultdiagnosis |