<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>This study addresses the fault diagnosis technology of CNC electromechanical systems in mechanical engineering equipment manufacturing, and explores the fault detection methods under the influence of structural coupling to improve the accuracy and efficiency of fault diagnosis. The study first analyzes the time-domain and frequency-domain features for fault diagnosis, including quantitative and dimensionless features used to identify different types of faults. Subsequently, the study explores feature dimensionality reduction methods, including algorithms such as PCA, LLE and t-SNE, and compares the effectiveness of their application in fault diagnosis. The research focuses on proposing a lightweight deep learning fault diagnosis framework called LTCN-BLS, which combines 2-DLTCN and 1-DLTCN branches, and an ILAEN-based BLS classifier to effectively extract and fuse time-domain and time-frequency-domain features of the data. The experimental results show that the LTCN-BLS framework has high accuracy and low network complexity in fault diagnosis, and has obvious advantages in early fault monitoring, degradation assessment, and robustness compared with traditional methods.</p>
</abstract>

This study addresses the fault diagnosis technology of CNC electromechanical systems in mechanical engineering equipment manufacturing, and explores the fault detection methods under the influence of structural coupling to improve the accuracy and efficiency of fault diagnosis. The study first analyzes the time-domain and frequency-domain features for fault diagnosis, including quantitative and dimensionless features used to identify different types of faults. Subsequently, the study explores feature dimensionality reduction methods, including algorithms such as PCA, LLE and t-SNE, and compares the effectiveness of their application in fault diagnosis. The research focuses on proposing a lightweight deep learning fault diagnosis framework called LTCN-BLS, which combines 2-DLTCN and 1-DLTCN branches, and an ILAEN-based BLS classifier to effectively extract and fuse time-domain and time-frequency-domain features of the data. The experimental results show that the LTCN-BLS framework has high accuracy and low network complexity in fault diagnosis, and has obvious advantages in early fault monitoring, degradation assessment, and robustness compared with traditional methods.

Fault diagnosis technology of CNC electromechanical system in mechanical engineering equipment manufacturing under structural coupling

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

This study addresses the fault diagnosis technology of CNC electromechanical systems in mechanical engineering equipment manufacturing, and explores the fault...

Fault diagnosis technology of CNC electromechanical system in mechanical engineering equipment manufacturing under structural coupling

Publicado en línea: 26 feb 2024

Páginas: -

Recibido: 17 ene 2024

Aceptado: 25 ene 2024

DOI: https://doi.org/10.2478/amns-2024-0688

Palabras clave
Numerical control electromechanical system, Fault diagnosis, Time domain, Frequency domain characterization

© 2024 Xueqing Bai, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Fault diagnosis technology of CNC electromechanical system in mechanical engineering equipment manufacturing under structural coupling

Publicado en línea: 26 feb 2024

Páginas: -

Recibido: 17 ene 2024

Aceptado: 25 ene 2024

DOI: https://doi.org/10.2478/amns-2024-0688

Palabras claveNumerical control electromechanical system, Fault diagnosis, Time domain, Frequency domain characterization

© 2024 Xueqing Bai, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Palabras clave
Numerical control electromechanical system, Fault diagnosis, Time domain, Frequency domain characterization