Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics

Kwangsuk Lee; Jae Kyeong Kim; Jaehyong Kim; Kyeon Hur; Hagbae Kim

doi:10.1109/ICKII.2018.8569065

Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics

Kwangsuk Lee, Jae Kyeong Kim, Jaehyong Kim, Kyeon Hur, Hagbae Kim

Department of Electrical and Electronic Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

12 Citations (Scopus)

Abstract

This paper proposes a multi-layered anomaly detection scheme to train feature extraction and to test anomaly prediction by using Convolutional Neural Networks (CNNs) layer, Bidirectional and Unidirectional Long Short-Term Memory (LSTM) Recurrent Neural Networks (RNNs), which is one of a novel deep architecture named stacked convolutional bidirectional LSTM network (SCB-LSTM). In the proposed model, the stacked CNNs perform feature extraction of vibration sensor signal patterns, and the result is used to feature learning with the stacked bidirectional LSTMs (SB-LSTMs). After this procedure, the stacked unidirectional LSTMs (SU-LSTMs) enhance the feature learning, and a regression layer finally predicts anomaly detections. The experimental results of bearing data not only show the accuracy of the proposed model in anomaly detection for rotating machinery diagnostics, but also suggest the better performance than other state-of-the-art algorithms such as a plain uni-LSTM or Bi-LSTM.

Original language	English
Title of host publication	1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018
Editors	Teen-Hang Meen
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	98-101
Number of pages	4
ISBN (Electronic)	9781538652671
DOIs	https://doi.org/10.1109/ICKII.2018.8569065
Publication status	Published - 2018 Dec 7
Event	1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018 - Jeju Island, Korea, Republic of Duration: 2018 Jul 23 → 2018 Jul 27

Publication series

Name	1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018

Other

Other	1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018
Country/Territory	Korea, Republic of
City	Jeju Island
Period	18/7/23 → 18/7/27

Bibliographical note

Publisher Copyright:
© 2018 IEEE.

All Science Journal Classification (ASJC) codes

Artificial Intelligence
Computer Graphics and Computer-Aided Design
Computer Networks and Communications
Computer Science Applications
Information Systems and Management
Media Technology

Access to Document

10.1109/ICKII.2018.8569065

Cite this

Lee, K., Kim, J. K., Kim, J., Hur, K., & Kim, H. (2018). Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics. In T-H. Meen (Ed.), 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018 (pp. 98-101). [8569065] (1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICKII.2018.8569065

Lee, Kwangsuk ; Kim, Jae Kyeong ; Kim, Jaehyong et al. / Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics. 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018. editor / Teen-Hang Meen. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 98-101 (1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018).

@inproceedings{6d11f18f09014fdaa55afdb5122a37c7,

title = "Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics",

abstract = "This paper proposes a multi-layered anomaly detection scheme to train feature extraction and to test anomaly prediction by using Convolutional Neural Networks (CNNs) layer, Bidirectional and Unidirectional Long Short-Term Memory (LSTM) Recurrent Neural Networks (RNNs), which is one of a novel deep architecture named stacked convolutional bidirectional LSTM network (SCB-LSTM). In the proposed model, the stacked CNNs perform feature extraction of vibration sensor signal patterns, and the result is used to feature learning with the stacked bidirectional LSTMs (SB-LSTMs). After this procedure, the stacked unidirectional LSTMs (SU-LSTMs) enhance the feature learning, and a regression layer finally predicts anomaly detections. The experimental results of bearing data not only show the accuracy of the proposed model in anomaly detection for rotating machinery diagnostics, but also suggest the better performance than other state-of-the-art algorithms such as a plain uni-LSTM or Bi-LSTM.",

author = "Kwangsuk Lee and Kim, {Jae Kyeong} and Jaehyong Kim and Kyeon Hur and Hagbae Kim",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018 ; Conference date: 23-07-2018 Through 27-07-2018",

year = "2018",

month = dec,

day = "7",

doi = "10.1109/ICKII.2018.8569065",

language = "English",

series = "1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "98--101",

editor = "Teen-Hang Meen",

booktitle = "1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018",

address = "United States",

}

Lee, K, Kim, JK, Kim, J, Hur, K & Kim, H 2018, Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics. in T-H Meen (ed.), 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018., 8569065, 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018, Institute of Electrical and Electronics Engineers Inc., pp. 98-101, 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018, Jeju Island, Korea, Republic of, 18/7/23. https://doi.org/10.1109/ICKII.2018.8569065

Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics. / Lee, Kwangsuk; Kim, Jae Kyeong; Kim, Jaehyong et al.

1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018. ed. / Teen-Hang Meen. Institute of Electrical and Electronics Engineers Inc., 2018. p. 98-101 8569065 (1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics

AU - Lee, Kwangsuk

AU - Kim, Jae Kyeong

AU - Kim, Jaehyong

AU - Hur, Kyeon

AU - Kim, Hagbae

PY - 2018/12/7

Y1 - 2018/12/7

N2 - This paper proposes a multi-layered anomaly detection scheme to train feature extraction and to test anomaly prediction by using Convolutional Neural Networks (CNNs) layer, Bidirectional and Unidirectional Long Short-Term Memory (LSTM) Recurrent Neural Networks (RNNs), which is one of a novel deep architecture named stacked convolutional bidirectional LSTM network (SCB-LSTM). In the proposed model, the stacked CNNs perform feature extraction of vibration sensor signal patterns, and the result is used to feature learning with the stacked bidirectional LSTMs (SB-LSTMs). After this procedure, the stacked unidirectional LSTMs (SU-LSTMs) enhance the feature learning, and a regression layer finally predicts anomaly detections. The experimental results of bearing data not only show the accuracy of the proposed model in anomaly detection for rotating machinery diagnostics, but also suggest the better performance than other state-of-the-art algorithms such as a plain uni-LSTM or Bi-LSTM.

AB - This paper proposes a multi-layered anomaly detection scheme to train feature extraction and to test anomaly prediction by using Convolutional Neural Networks (CNNs) layer, Bidirectional and Unidirectional Long Short-Term Memory (LSTM) Recurrent Neural Networks (RNNs), which is one of a novel deep architecture named stacked convolutional bidirectional LSTM network (SCB-LSTM). In the proposed model, the stacked CNNs perform feature extraction of vibration sensor signal patterns, and the result is used to feature learning with the stacked bidirectional LSTMs (SB-LSTMs). After this procedure, the stacked unidirectional LSTMs (SU-LSTMs) enhance the feature learning, and a regression layer finally predicts anomaly detections. The experimental results of bearing data not only show the accuracy of the proposed model in anomaly detection for rotating machinery diagnostics, but also suggest the better performance than other state-of-the-art algorithms such as a plain uni-LSTM or Bi-LSTM.

UR - http://www.scopus.com/inward/record.url?scp=85060724201&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85060724201&partnerID=8YFLogxK

U2 - 10.1109/ICKII.2018.8569065

DO - 10.1109/ICKII.2018.8569065

M3 - Conference contribution

AN - SCOPUS:85060724201

T3 - 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018

SP - 98

EP - 101

BT - 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018

A2 - Meen, Teen-Hang

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018

Y2 - 23 July 2018 through 27 July 2018

ER -

Lee K, Kim JK, Kim J, Hur K, Kim H. Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics. In Meen T-H, editor, 1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 98-101. 8569065. (1st IEEE International Conference on Knowledge Innovation and Invention, ICKII 2018). doi: 10.1109/ICKII.2018.8569065

Stacked Convolutional Bidirectional LSTM Recurrent Neural Network for Bearing Anomaly Detection in Rotating Machinery Diagnostics

Abstract

Publication series

Other

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this