Cross-device fault diagnosis method based on graph convolution and multi-sensor fusion
ObjectiveFor mechanical equipment in actual production, it is difficult or impossible to obtain a large amount of labeled data, resulting in low accuracy of traditional fault diagnosis methods. To address this problem, a cross-device fault diagnosis method based on graph convolution and multi-sensor...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Editorial Office of Journal of Mechanical Strength
2024-01-01
|
| Series: | Jixie qiangdu |
| Subjects: | |
| Online Access: | http://www.jxqd.net.cn/thesisDetails?columnId=79651143&Fpath=home&index=0 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841534052323033088 |
|---|---|
| author | SUN Yuanshuai KONG Fanqin NIE Xiaoyin XIE Gang |
| author_facet | SUN Yuanshuai KONG Fanqin NIE Xiaoyin XIE Gang |
| author_sort | SUN Yuanshuai |
| collection | DOAJ |
| description | ObjectiveFor mechanical equipment in actual production, it is difficult or impossible to obtain a large amount of labeled data, resulting in low accuracy of traditional fault diagnosis methods. To address this problem, a cross-device fault diagnosis method based on graph convolution and multi-sensor fusion, convolutional domain graph convolution network (CDGCN) , was proposed. This method can model class labels, domain labels, and data feature structures.MethodsFirstly, a convolutional neural network was used to extract features from the input signal. Then, the feature structure relationship of the sample was mined through the graph generation layer to construct an instance graph. The instance graph was modeled using a graph convolutional neural network, and a multi-sensor high-level feature fusion method was proposed to perform multi-sensor information fusion. Finally, domain adaptation was achieved by using distribution difference metrics, classifiers, and domain discriminators.ResultsThe proposed method can capture domain-invariant features and discriminant features, and ultimately achieve cross-device fault diagnosis. Migration experiments on two datasets show that the proposed CDGCN not only achieves the best performance among the compared methods, but also extracts transferable features for cross-device domain adaptation. |
| format | Article |
| id | doaj-art-039d46e0a7d34bf2b51e188ee7312f8b |
| institution | Kabale University |
| issn | 1001-9669 |
| language | zho |
| publishDate | 2024-01-01 |
| publisher | Editorial Office of Journal of Mechanical Strength |
| record_format | Article |
| series | Jixie qiangdu |
| spelling | doaj-art-039d46e0a7d34bf2b51e188ee7312f8b2025-01-15T02:46:09ZzhoEditorial Office of Journal of Mechanical StrengthJixie qiangdu1001-96692024-01-0111079651143Cross-device fault diagnosis method based on graph convolution and multi-sensor fusionSUN YuanshuaiKONG FanqinNIE XiaoyinXIE GangObjectiveFor mechanical equipment in actual production, it is difficult or impossible to obtain a large amount of labeled data, resulting in low accuracy of traditional fault diagnosis methods. To address this problem, a cross-device fault diagnosis method based on graph convolution and multi-sensor fusion, convolutional domain graph convolution network (CDGCN) , was proposed. This method can model class labels, domain labels, and data feature structures.MethodsFirstly, a convolutional neural network was used to extract features from the input signal. Then, the feature structure relationship of the sample was mined through the graph generation layer to construct an instance graph. The instance graph was modeled using a graph convolutional neural network, and a multi-sensor high-level feature fusion method was proposed to perform multi-sensor information fusion. Finally, domain adaptation was achieved by using distribution difference metrics, classifiers, and domain discriminators.ResultsThe proposed method can capture domain-invariant features and discriminant features, and ultimately achieve cross-device fault diagnosis. Migration experiments on two datasets show that the proposed CDGCN not only achieves the best performance among the compared methods, but also extracts transferable features for cross-device domain adaptation.http://www.jxqd.net.cn/thesisDetails?columnId=79651143&Fpath=home&index=0Graph convolutional neural networkMulti-sensorCross-deviceDomain adaptationFault diagnosis |
| spellingShingle | SUN Yuanshuai KONG Fanqin NIE Xiaoyin XIE Gang Cross-device fault diagnosis method based on graph convolution and multi-sensor fusion Jixie qiangdu Graph convolutional neural network Multi-sensor Cross-device Domain adaptation Fault diagnosis |
| title | Cross-device fault diagnosis method based on graph convolution and multi-sensor fusion |
| title_full | Cross-device fault diagnosis method based on graph convolution and multi-sensor fusion |
| title_fullStr | Cross-device fault diagnosis method based on graph convolution and multi-sensor fusion |
| title_full_unstemmed | Cross-device fault diagnosis method based on graph convolution and multi-sensor fusion |
| title_short | Cross-device fault diagnosis method based on graph convolution and multi-sensor fusion |
| title_sort | cross device fault diagnosis method based on graph convolution and multi sensor fusion |
| topic | Graph convolutional neural network Multi-sensor Cross-device Domain adaptation Fault diagnosis |
| url | http://www.jxqd.net.cn/thesisDetails?columnId=79651143&Fpath=home&index=0 |
| work_keys_str_mv | AT sunyuanshuai crossdevicefaultdiagnosismethodbasedongraphconvolutionandmultisensorfusion AT kongfanqin crossdevicefaultdiagnosismethodbasedongraphconvolutionandmultisensorfusion AT niexiaoyin crossdevicefaultdiagnosismethodbasedongraphconvolutionandmultisensorfusion AT xiegang crossdevicefaultdiagnosismethodbasedongraphconvolutionandmultisensorfusion |