| 作成者 |
|
|
|
|
|
|
|
|
|
| 本文言語 |
|
| 出版者 |
|
|
|
| 発行日 |
|
| 収録物名 |
|
| 巻 |
|
| 号 |
|
| 開始ページ |
|
| 終了ページ |
|
| 出版タイプ |
|
| アクセス権 |
|
| Crossref DOI |
|
| 権利関係 |
|
| 概要 |
In Industry 4.0, condition monitoring is gaining significant importance due to its capability for early diagnosis of the symptoms of the faults that result in severe damage in the rotating machinery. ...Hydrodynamic conical journal bearings (HCJB) are used in heavy rotating machinery due to their high radial and axial load-carrying capacity and need to be adequately monitored during the operation. Thus, an experimental study is carried out on a healthy and faulty CJB which is scratched due to frequent start and stop operation, i.e. worn-out internal surfaces of bearing. Therefore, the dynamic response of the journal changes significantly due to wear. The vibration response of bearing in the time domain was acquired and transformed into the frequency domain and a comparative analysis is carried out between the bearings. The experimental results demonstrate the unique vibration characteristic of worn-out bearing at the 4th multiple of operating frequency. Machine learning techniques have also been used as fault diagnosis approaches. A comparative study using machine learning methods i.e. SVM (Support Vector Machine), k-NN (k-Nearest Neighbour) and RF (Random Forest) has been performed. The results show that the RF classifier is more proficient in fault diagnosis with an accuracy of 93.93% compared to the other methods i.e. k-NN and SVM.続きを見る
|
Mendeley出力
