| Autor: |
Wahul, Revati M., Kale, Archana P., Patange, Abhishek D. |
| Předmět: |
|
| Zdroj: |
International Journal of Performability Engineering; Oct2022, Vol. 18 Issue 10, p730-740, 11p |
| Abstrakt: |
In consideration of high-precision machining, Tool Condition Monitoring (TCM) is very influential to retain the resolution and accuracy of the machined component. TCM is considered to be always challenging due to diversified operating conditions. In an era of big data analytics, Deep Learning-based networks are gaining significant consideration in the manufacturing industry in the direction of dealing with data gathered from these diversified conditions in a heavy noise environment. In order to address these problems, the Convolutional Neural Network (CNN) based deep learning approach is advocated herein for condition monitoring of a turning tool. After acquiring real-time vibration data corresponding to tool faults, the CNN architecture was designed to assign decimal probabilities to every category in a multi-class classification of tool faults followed by hyperparameters tuning. A rigorous analysis was undertaken through different datasets gathered from diversified machining conditions. The test and validation results demonstrated that the proposed network outperforms the conventional machine learning classifiers. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
