An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM

5494 Views
5595 Downloads
Export citation: ABNT
PAN, Honghu ;HE, Xingxi ;TANG, Sai ;MENG, Fanming .
An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM. 
Strojniški vestnik - Journal of Mechanical Engineering, [S.l.], v. 64, n.7-8, p. 443-452, july 2018. 
ISSN 0039-2480.
Available at: <https://www.sv-jme.eu/article/an-improved-bearing-fault-diagnosis-method-using-one-dimensional-cnn-and-lstm/>. Date accessed: 29 mar. 2024. 
doi:http://dx.doi.org/10.5545/sv-jme.2018.5249.
Pan, H., He, X., Tang, S., & Meng, F.
(2018).
An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM.
Strojniški vestnik - Journal of Mechanical Engineering, 64(7-8), 443-452.
doi:http://dx.doi.org/10.5545/sv-jme.2018.5249
@article{sv-jmesv-jme.2018.5249,
	author = {Honghu  Pan and Xingxi  He and Sai  Tang and Fanming  Meng},
	title = {An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM},
	journal = {Strojniški vestnik - Journal of Mechanical Engineering},
	volume = {64},
	number = {7-8},
	year = {2018},
	keywords = {bearing fault diagnosis; CNN; LSTM},
	abstract = {As one of the most critical components in rotating machinery, bearing fault diagnosis has attracted many researchers’ attention. The traditional methods for bearing fault diagnosis normally requires three steps, including data pre-processing, feature extraction and pattern classification, which require much expertise and experience. This paper takes advantage of deep learning algorithms and proposes an improved bearing fault diagnosis method based on a convolutional neural network (CNN) and a long-short-term memory (LSTM) recurrent neural network whose input is the raw sampling signal without any pre-processing or traditional feature extraction. The CNN is frequently used in image classification as it could extract features automatically from high-dimensional data, while LSTM is most applied in speech recognition as it considers time coherence. This paper combined one-dimensional CNN and LSTM into one unified structure by using the CNN’s output as input to the LSTM to identify the bearing fault types. First, a part of raw bearing signal data is used as the training dataset in the model, and the simulation ends when the number of iterations reaches a specific value. Second, the rest of the signal data was input in the trained model as the testing dataset to verify the effectiveness of the proposed method. The results show that the average accuracy rate in the testing dataset of this proposed method reaches more than 99 %, which outperforms other algorithms for bearing fault diagnosis.},
	issn = {0039-2480},	pages = {443-452},	doi = {10.5545/sv-jme.2018.5249},
	url = {https://www.sv-jme.eu/article/an-improved-bearing-fault-diagnosis-method-using-one-dimensional-cnn-and-lstm/}
}
Pan, H.,He, X.,Tang, S.,Meng, F.
2018 July 64. An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM. Strojniški vestnik - Journal of Mechanical Engineering. [Online] 64:7-8
%A Pan, Honghu 
%A He, Xingxi 
%A Tang, Sai 
%A Meng, Fanming 
%D 2018
%T An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM
%B 2018
%9 bearing fault diagnosis; CNN; LSTM
%! An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM
%K bearing fault diagnosis; CNN; LSTM
%X As one of the most critical components in rotating machinery, bearing fault diagnosis has attracted many researchers’ attention. The traditional methods for bearing fault diagnosis normally requires three steps, including data pre-processing, feature extraction and pattern classification, which require much expertise and experience. This paper takes advantage of deep learning algorithms and proposes an improved bearing fault diagnosis method based on a convolutional neural network (CNN) and a long-short-term memory (LSTM) recurrent neural network whose input is the raw sampling signal without any pre-processing or traditional feature extraction. The CNN is frequently used in image classification as it could extract features automatically from high-dimensional data, while LSTM is most applied in speech recognition as it considers time coherence. This paper combined one-dimensional CNN and LSTM into one unified structure by using the CNN’s output as input to the LSTM to identify the bearing fault types. First, a part of raw bearing signal data is used as the training dataset in the model, and the simulation ends when the number of iterations reaches a specific value. Second, the rest of the signal data was input in the trained model as the testing dataset to verify the effectiveness of the proposed method. The results show that the average accuracy rate in the testing dataset of this proposed method reaches more than 99 %, which outperforms other algorithms for bearing fault diagnosis.
%U https://www.sv-jme.eu/article/an-improved-bearing-fault-diagnosis-method-using-one-dimensional-cnn-and-lstm/
%0 Journal Article
%R 10.5545/sv-jme.2018.5249
%& 443
%P 10
%J Strojniški vestnik - Journal of Mechanical Engineering
%V 64
%N 7-8
%@ 0039-2480
%8 2018-07-12
%7 2018-07-12
Pan, Honghu, Xingxi  He, Sai  Tang, & Fanming  Meng.
"An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM." Strojniški vestnik - Journal of Mechanical Engineering [Online], 64.7-8 (2018): 443-452. Web.  29 Mar. 2024
TY  - JOUR
AU  - Pan, Honghu 
AU  - He, Xingxi 
AU  - Tang, Sai 
AU  - Meng, Fanming 
PY  - 2018
TI  - An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM
JF  - Strojniški vestnik - Journal of Mechanical Engineering
DO  - 10.5545/sv-jme.2018.5249
KW  - bearing fault diagnosis; CNN; LSTM
N2  - As one of the most critical components in rotating machinery, bearing fault diagnosis has attracted many researchers’ attention. The traditional methods for bearing fault diagnosis normally requires three steps, including data pre-processing, feature extraction and pattern classification, which require much expertise and experience. This paper takes advantage of deep learning algorithms and proposes an improved bearing fault diagnosis method based on a convolutional neural network (CNN) and a long-short-term memory (LSTM) recurrent neural network whose input is the raw sampling signal without any pre-processing or traditional feature extraction. The CNN is frequently used in image classification as it could extract features automatically from high-dimensional data, while LSTM is most applied in speech recognition as it considers time coherence. This paper combined one-dimensional CNN and LSTM into one unified structure by using the CNN’s output as input to the LSTM to identify the bearing fault types. First, a part of raw bearing signal data is used as the training dataset in the model, and the simulation ends when the number of iterations reaches a specific value. Second, the rest of the signal data was input in the trained model as the testing dataset to verify the effectiveness of the proposed method. The results show that the average accuracy rate in the testing dataset of this proposed method reaches more than 99 %, which outperforms other algorithms for bearing fault diagnosis.
UR  - https://www.sv-jme.eu/article/an-improved-bearing-fault-diagnosis-method-using-one-dimensional-cnn-and-lstm/
@article{{sv-jme}{sv-jme.2018.5249},
	author = {Pan, H., He, X., Tang, S., Meng, F.},
	title = {An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM},
	journal = {Strojniški vestnik - Journal of Mechanical Engineering},
	volume = {64},
	number = {7-8},
	year = {2018},
	doi = {10.5545/sv-jme.2018.5249},
	url = {https://www.sv-jme.eu/article/an-improved-bearing-fault-diagnosis-method-using-one-dimensional-cnn-and-lstm/}
}
TY  - JOUR
AU  - Pan, Honghu 
AU  - He, Xingxi 
AU  - Tang, Sai 
AU  - Meng, Fanming 
PY  - 2018/07/12
TI  - An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM
JF  - Strojniški vestnik - Journal of Mechanical Engineering; Vol 64, No 7-8 (2018): Strojniški vestnik - Journal of Mechanical Engineering
DO  - 10.5545/sv-jme.2018.5249
KW  - bearing fault diagnosis, CNN, LSTM
N2  - As one of the most critical components in rotating machinery, bearing fault diagnosis has attracted many researchers’ attention. The traditional methods for bearing fault diagnosis normally requires three steps, including data pre-processing, feature extraction and pattern classification, which require much expertise and experience. This paper takes advantage of deep learning algorithms and proposes an improved bearing fault diagnosis method based on a convolutional neural network (CNN) and a long-short-term memory (LSTM) recurrent neural network whose input is the raw sampling signal without any pre-processing or traditional feature extraction. The CNN is frequently used in image classification as it could extract features automatically from high-dimensional data, while LSTM is most applied in speech recognition as it considers time coherence. This paper combined one-dimensional CNN and LSTM into one unified structure by using the CNN’s output as input to the LSTM to identify the bearing fault types. First, a part of raw bearing signal data is used as the training dataset in the model, and the simulation ends when the number of iterations reaches a specific value. Second, the rest of the signal data was input in the trained model as the testing dataset to verify the effectiveness of the proposed method. The results show that the average accuracy rate in the testing dataset of this proposed method reaches more than 99 %, which outperforms other algorithms for bearing fault diagnosis.
UR  - https://www.sv-jme.eu/article/an-improved-bearing-fault-diagnosis-method-using-one-dimensional-cnn-and-lstm/
Pan, Honghu, He, Xingxi, Tang, Sai, AND Meng, Fanming.
"An Improved Bearing Fault Diagnosis Method using One-Dimensional CNN and LSTM" Strojniški vestnik - Journal of Mechanical Engineering [Online], Volume 64 Number 7-8 (12 July 2018)

Authors

Affiliations

  • Chongqing University, The State Key Laboratory of Mechanical Transmissions, China 1

Paper's information

Strojniški vestnik - Journal of Mechanical Engineering 64(2018)7-8, 443-452
© The Authors, CC-BY 4.0 Int. Change in copyright policy from 2022, Jan 1st.

https://doi.org/10.5545/sv-jme.2018.5249

As one of the most critical components in rotating machinery, bearing fault diagnosis has attracted many researchers’ attention. The traditional methods for bearing fault diagnosis normally requires three steps, including data pre-processing, feature extraction and pattern classification, which require much expertise and experience. This paper takes advantage of deep learning algorithms and proposes an improved bearing fault diagnosis method based on a convolutional neural network (CNN) and a long-short-term memory (LSTM) recurrent neural network whose input is the raw sampling signal without any pre-processing or traditional feature extraction. The CNN is frequently used in image classification as it could extract features automatically from high-dimensional data, while LSTM is most applied in speech recognition as it considers time coherence. This paper combined one-dimensional CNN and LSTM into one unified structure by using the CNN’s output as input to the LSTM to identify the bearing fault types. First, a part of raw bearing signal data is used as the training dataset in the model, and the simulation ends when the number of iterations reaches a specific value. Second, the rest of the signal data was input in the trained model as the testing dataset to verify the effectiveness of the proposed method. The results show that the average accuracy rate in the testing dataset of this proposed method reaches more than 99 %, which outperforms other algorithms for bearing fault diagnosis.

bearing fault diagnosis; CNN; LSTM