期刊
JOURNAL OF MANUFACTURING SYSTEMS
卷 59, 期 -, 页码 467-480出版社
ELSEVIER SCI LTD
DOI: 10.1016/j.jmsy.2021.03.022
关键词
Bearing fault diagnosis; Convolutional neural network; Feature learning; Adaptive kernel width; Dynamic receptive field; Sparse regularization
资金
- National Natural Science Foundation of China [71777173]
- Equipment advance research foundation [61400020119]
- Fundamental Research Funds for the Central Universities [22120190196]
- Science and Technology Innovation Action Plan of Shanghai Science and Technology Commission [19511106303]
The paper proposes a new deep neural network (DNN) model, adaptive kernel sparse network (AKSNet), to extract multi-scale fault features from vibration signals. AKSNet integrates key techniques such as adaptive kernel selection, channel attention, and spatial attention to effectively improve fault diagnosis performance of the classifier.
Convolutional kernels have significant affections on feature learning of convolutional neural network (CNN). However, it is still a challenging problem to determine appropriate kernel width. Moreover, some features learned by convolutional layers are still redundant and noisy. Thus, adaptive selection of kernel width and feature selection of feature maps are key techniques to improve feature learning performance of CNNs. In this paper, a new deep neural network (DNN) model, adaptive kernel sparse network (AKSNet) is proposed to extract multi-scale fault features from one-dimensional (1-D) vibration signals. Firstly, an adaptive kernel selection method is developed, where multiple branches with different kernels are used to extract multi-scale features from vibration signals. Channel-wise attention is developed to fuse features generated by these kernels to obtain different informative scales. Secondly, a spatial attention is used for dynamic receptive field to focus on salient region of feature maps. Thirdly, a sparse regularization layer is embedded in the deep network to further filter noise and highlight impaction of the feature maps. Finally, two cases are adopted to verify effectiveness of AKSNet-based feature learning for bearing fault diagnosis. Experimental results show that AKSNet can effectively extract features from multi-channel vibration signals and then improves fault diagnosis performance of the classifier significantly. AKSNet shows better recognition performance in comparison with that of shallow neural networks and other typical DNNs.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据