Rolling Bearing Fault Image Recognition Method Based On GWO Optimization Darknet19

Failure of rolling bearings is a common mechanical problem, and traditional fault diagnosis methods require experienced experts and have limited effectiveness for large-scale data processing and analysis. This study optimizes the parameters of a rolling fault image recognition model for Darknet19 ne...

Full description

Saved in:
Bibliographic Details
Published in2024 2nd International Conference on Computer Graphics and Image Processing (CGIP) pp. 38 - 42
Main Authors Zhang, Jinrong, Liang, Jingxin, Shang, Sitong
Format Conference Proceeding
LanguageEnglish
Published IEEE 12.01.2024
Subjects
Darknet19
Fault diagnosis
fault image recognition
GWO
Hyperparameter optimization
Image recognition
Production
rolling bearing
Rolling bearings
Training
Vibrations
Online AccessGet full text

Cover

More Information
Summary:Failure of rolling bearings is a common mechanical problem, and traditional fault diagnosis methods require experienced experts and have limited effectiveness for large-scale data processing and analysis. This study optimizes the parameters of a rolling fault image recognition model for Darknet19 network based on the Gray Wolf Optimization algorithm. The model adjusts the structure and parameters of the network by comparing different neural networks horizontally to make it more suitable for rolling bearing fault diagnosis tasks. Experimental results show that the Darknet19 network with optimized parameters by the GWO algorithm achieves significant performance improvement in rolling bearing fault diagnosis. The optimized model exhibits higher diagnostic accuracy and robustness under multiple fault categories. Therefore, the rolling bearing fault image recognition method based on GWO and Darknet19 can provide efficient and accurate automatic fault diagnosis.
DOI:10.1109/CGIP62525.2024.00015