CNN4GCDD : a one-dimensional convolutional neural network-based model for gear crack depth diagnosis

Zhang, Shouhua; Zhou, Jiehan; Wang, Erhua; Pirttikangas, Susanna (2022-05-20)
 
Avaa tiedosto
Lataukset: 

URL:
https://doi.org/10.1109/cscwd54268.2022.9776142

Zhang, Shouhua
Zhou, Jiehan
Wang, Erhua
Pirttikangas, Susanna
Institute of Electrical and Electronics Engineers
20.05.2022

S. Zhang, J. Zhou, E. Wang and S. Pirttikangas, "CNN4GCDD: a One-Dimensional Convolutional Neural Network-based Model for Gear Crack Depth Diagnosis," 2022 IEEE 25th International Conference on Computer Supported Cooperative Work in Design (CSCWD), 2022, pp. 1138-1142, doi: 10.1109/CSCWD54268.2022.9776142.

https://rightsstatements.org/vocab/InC/1.0/
© 2022 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
https://rightsstatements.org/vocab/InC/1.0/
doi:https://doi.org/10.1109/cscwd54268.2022.9776142
Näytä kaikki kuvailutiedot
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2022101061477
Tiivistelmä

Abstract

Gear crack is one of the common failures in transmission systems. With the gradual expansion of cracks, it may cause tooth fracture. Therefore, it is of great significance to study the fault diagnosis of gear cracks. Vibration signals with time sequence are widely used in gear fault diagnosis. Extracting key fault features from vibration signals determines the accuracy of fault diagnosis models. This paper takes spur gears as research objects, and proposes a model for diagnosing gear crack depth based on one-dimensional convolutional neural network (short for CNN4GCDD). In order to identify crack depths, we collect the vibration signals from three gears with various crack depths and a normal gear without cracks. CNN4GCDD uses the original vibration signal as the input, adaptively extracts features, and makes crack depth diagnosis through the convolutional neural network. The experimental results demonstrate that CNN4GCDD can directly use the original time-domain signal for crack depth diagnosis, and make a high accurate prediction.

Kokoelmat