A lightweight and precision dual track 1D and 2D feature fusion convolutional network for machinery equipment fault diagnosis
Abstract Addressing the issues of a single-feature input channel structure, scarcity of training fault data, and insufficient feature learning capabilities in noisy environments for intelligent diagnostic models of mechanical equipment, we propose a method based on a one-dimensional and two-dimensio...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-12-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-81118-2 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841559532179816448 |
|---|---|
| author | Chaoquan Mo Ke Huang Houxin Ji |
| author_facet | Chaoquan Mo Ke Huang Houxin Ji |
| author_sort | Chaoquan Mo |
| collection | DOAJ |
| description | Abstract Addressing the issues of a single-feature input channel structure, scarcity of training fault data, and insufficient feature learning capabilities in noisy environments for intelligent diagnostic models of mechanical equipment, we propose a method based on a one-dimensional and two-dimensional dual-channel feature information fusion convolutional neural network (1D_2DIFCNN). By constructing a one-dimensional and two-dimensiona dual-channel feature information fusion convolutional network and introducing a Convolutional Block Attention Mechanism, we utilize Random Overlapping Sampling Technique to process raw vibration signals. The model takes as inputs both one-dimensional data and two-dimensional Continuous Wavelet Transform images. Experimental validation shows that this method exhibits faster convergence, higher diagnostic accuracy, and good robustness and generalization performance on two different datasets, outperforming other advanced algorithms. |
| format | Article |
| id | doaj-art-21c19fce9b6c43a7b932ba74a3cbfaf5 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-21c19fce9b6c43a7b932ba74a3cbfaf52025-01-05T12:25:03ZengNature PortfolioScientific Reports2045-23222024-12-0114111910.1038/s41598-024-81118-2A lightweight and precision dual track 1D and 2D feature fusion convolutional network for machinery equipment fault diagnosisChaoquan Mo0Ke Huang1Houxin Ji2College of Mechanical and Electrical Engineering, Wenzhou UniversityCollege of Mechanical and Electrical Engineering, Wenzhou UniversityCollege of Mechanical and Electrical Engineering, Wenzhou UniversityAbstract Addressing the issues of a single-feature input channel structure, scarcity of training fault data, and insufficient feature learning capabilities in noisy environments for intelligent diagnostic models of mechanical equipment, we propose a method based on a one-dimensional and two-dimensional dual-channel feature information fusion convolutional neural network (1D_2DIFCNN). By constructing a one-dimensional and two-dimensiona dual-channel feature information fusion convolutional network and introducing a Convolutional Block Attention Mechanism, we utilize Random Overlapping Sampling Technique to process raw vibration signals. The model takes as inputs both one-dimensional data and two-dimensional Continuous Wavelet Transform images. Experimental validation shows that this method exhibits faster convergence, higher diagnostic accuracy, and good robustness and generalization performance on two different datasets, outperforming other advanced algorithms.https://doi.org/10.1038/s41598-024-81118-2Mechanical equipment fault diagnosis1D-2D dual-channel information fusionConvolutional attention mechanismContinuous Wavelet TransformFeature fusion |
| spellingShingle | Chaoquan Mo Ke Huang Houxin Ji A lightweight and precision dual track 1D and 2D feature fusion convolutional network for machinery equipment fault diagnosis Scientific Reports Mechanical equipment fault diagnosis 1D-2D dual-channel information fusion Convolutional attention mechanism Continuous Wavelet Transform Feature fusion |
| title | A lightweight and precision dual track 1D and 2D feature fusion convolutional network for machinery equipment fault diagnosis |
| title_full | A lightweight and precision dual track 1D and 2D feature fusion convolutional network for machinery equipment fault diagnosis |
| title_fullStr | A lightweight and precision dual track 1D and 2D feature fusion convolutional network for machinery equipment fault diagnosis |
| title_full_unstemmed | A lightweight and precision dual track 1D and 2D feature fusion convolutional network for machinery equipment fault diagnosis |
| title_short | A lightweight and precision dual track 1D and 2D feature fusion convolutional network for machinery equipment fault diagnosis |
| title_sort | lightweight and precision dual track 1d and 2d feature fusion convolutional network for machinery equipment fault diagnosis |
| topic | Mechanical equipment fault diagnosis 1D-2D dual-channel information fusion Convolutional attention mechanism Continuous Wavelet Transform Feature fusion |
| url | https://doi.org/10.1038/s41598-024-81118-2 |
| work_keys_str_mv | AT chaoquanmo alightweightandprecisiondualtrack1dand2dfeaturefusionconvolutionalnetworkformachineryequipmentfaultdiagnosis AT kehuang alightweightandprecisiondualtrack1dand2dfeaturefusionconvolutionalnetworkformachineryequipmentfaultdiagnosis AT houxinji alightweightandprecisiondualtrack1dand2dfeaturefusionconvolutionalnetworkformachineryequipmentfaultdiagnosis AT chaoquanmo lightweightandprecisiondualtrack1dand2dfeaturefusionconvolutionalnetworkformachineryequipmentfaultdiagnosis AT kehuang lightweightandprecisiondualtrack1dand2dfeaturefusionconvolutionalnetworkformachineryequipmentfaultdiagnosis AT houxinji lightweightandprecisiondualtrack1dand2dfeaturefusionconvolutionalnetworkformachineryequipmentfaultdiagnosis |