Rotating machinery fault diagnosis based on a novel lightweight convolutional neural network

Jing Yan, Tingliang Liu, Xinyu Ye, Qianzhen Jing and Yuannan Dai

PLOS ONE, 2021, vol. 16, issue 8, 1-20

Abstract: The advancement of Industry 4.0 and Industrial Internet of Things (IIoT) has laid more emphasis on reducing the parameter amount and storage space of the model in addition to the automatic and accurate fault diagnosis. In this case, this paper proposes a lightweight convolutional neural network (LCNN) method for intelligent fault diagnosis of rotating machinery, which can largely satisfy the need of less parameter amount and storage space as well as high accuracy. First, light-weight convolution blocks are constructed through basic elements such as spatial separable convolutions with the aim to effectively reduce model parameters. Secondly, the LCNN model for the intelligent fault diagnosis is constructed via lightweight convolution blocks instead of the tradi-tional convolution operation. Finally, to address the “black box” problem, the entire network is visualized through Tensorboard and t-distribution stochastic neighbor embedding. The results demonstrate that when the number of lightweight convolutional blocks reaches 6, the diagnosis accuracy of the LCNN model exceeds 99.9%. And the proposed model has become the most robust with parameters significantly decreasing. Furthermore, the proposed LCNN model has realized accurate, automatic, and robust fault diagnosis of rotating machinery, which makes it more suitable for deployment under the IIoT context.

Date: 2021
References: View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0256287 (text/html)
https://journals.plos.org/plosone/article/file?id= ... 56287&type=printable (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:plo:pone00:0256287

DOI: 10.1371/journal.pone.0256287

Access Statistics for this article

More articles in PLOS ONE from Public Library of Science
Bibliographic data for series maintained by plosone ().