2021
DOI: 10.1016/j.measurement.2020.108868
|View full text |Cite
|
Sign up to set email alerts
|

Bearing performance degradation assessment based on optimized EWT and CNN

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

0
12
0

Year Published

2021
2021
2023
2023

Publication Types

Select...
8
1

Relationship

0
9

Authors

Journals

citations
Cited by 33 publications
(12 citation statements)
references
References 21 publications
0
12
0
Order By: Relevance
“…Mao et al (2020) have proposed a Safe Semi-Supervised Support Vector Machine (S4VM) algorithm, which detects that the early abnormal point of bearing 1 is the sampling sequence 535. Hu et al (2020) have introduced the degradation index constructed based on Empirical Wavelet Transform (EWT) and CNN, which can detect that bearing 1 enters the early anomaly when the serial number is 533. These results are basically the same as the method proposed in this paper.…”
Section: Resultsmentioning
confidence: 99%
See 1 more Smart Citation
“…Mao et al (2020) have proposed a Safe Semi-Supervised Support Vector Machine (S4VM) algorithm, which detects that the early abnormal point of bearing 1 is the sampling sequence 535. Hu et al (2020) have introduced the degradation index constructed based on Empirical Wavelet Transform (EWT) and CNN, which can detect that bearing 1 enters the early anomaly when the serial number is 533. These results are basically the same as the method proposed in this paper.…”
Section: Resultsmentioning
confidence: 99%
“…Singular values are used as input to construct a graph, and GMSVs are used to check whether the rolling bearing has a failure during continuous operation, which improves the sensitivity to early failure and the robustness of index. Hu et al (2020) have used the Convolutional Neural Network (CNN) to extract the sensitive features of the Short-Time Fourier Transform (STFT) envelope spectrum in bearing fault component, which can improve the sensitivity and stability of the early anomaly detection. The above methods are suitable for the case where the monitoring data are small, and when the sample size is large, problems such as low efficiency and large calculation amount occur.…”
Section: Introductionmentioning
confidence: 99%
“…However, traditional machine learning methods like SVM require a priori knowledge of feature engineering, which is extremely difficult to implement with regard to bearings due to the complex working conditions they operate under. Deep learning-based algorithms provide an alternative solution to this problem [ 8 , 9 , 10 ]. Chen et al proposed a method based on neuro-fuzzy systems (NFSs) and Bayesian algorithms, which use trained NFSs as predictors to discern the degradation of a given machine’s fault state over time [ 11 ].…”
Section: Introductionmentioning
confidence: 99%
“…The existing popular empirical mode decomposition [24] often has the problem of modal aliasing. Furthermore, the difficulty of wavelet decomposition [25] lies in how to effectively select the wavelet basis and decomposition scale. In addition, the decomposition methods may introduce some redundant decomposition information to the predictive models, degenerating the predictive computational cost.…”
Section: Introductionmentioning
confidence: 99%