Fault Detection and Diagnosis Using Combined Autoencoder and Long Short-Term Memory Network
Fault detection and diagnosis is one of the most critical components of preventing accidents and ensuring the system safety of industrial processes. In this paper, we propose an integrated learning approach for jointly achieving fault detection and fault diagnosis of rare events in multivariate time...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-10-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/19/21/4612 |
| _version_ | 1811262351845883904 |
|---|---|
| author | Pangun Park Piergiuseppe Di Marco Hyejeon Shin Junseong Bang |
| author_facet | Pangun Park Piergiuseppe Di Marco Hyejeon Shin Junseong Bang |
| author_sort | Pangun Park |
| collection | DOAJ |
| description | Fault detection and diagnosis is one of the most critical components of preventing accidents and ensuring the system safety of industrial processes. In this paper, we propose an integrated learning approach for jointly achieving fault detection and fault diagnosis of rare events in multivariate time series data. The proposed approach combines an autoencoder to detect a rare fault event and a long short-term memory (LSTM) network to classify different types of faults. The autoencoder is trained with offline normal data, which is then used as the anomaly detection. The predicted faulty data, captured by autoencoder, are put into the LSTM network to identify the types of faults. It basically combines the strong low-dimensional nonlinear representations of the autoencoder for the rare event detection and the strong time series learning ability of LSTM for the fault diagnosis. The proposed approach is compared with a deep convolutional neural network approach for fault detection and identification on the Tennessee Eastman process. Experimental results show that the combined approach accurately detects deviations from normal behaviour and identifies the types of faults within the useful time. |
| first_indexed | 2024-04-12T19:23:48Z |
| format | Article |
| id | doaj.art-54525d449ef049ab98812a95ff16da61 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-12T19:23:48Z |
| publishDate | 2019-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-54525d449ef049ab98812a95ff16da612022-12-22T03:19:32ZengMDPI AGSensors1424-82202019-10-011921461210.3390/s19214612s19214612Fault Detection and Diagnosis Using Combined Autoencoder and Long Short-Term Memory NetworkPangun Park0Piergiuseppe Di Marco1Hyejeon Shin2Junseong Bang3Department of Radio and Information Communications Engineering, Chungnam National University, Daejeon 34134, KoreaDepartment of Information Engineering, Computer Science and Mathematics, University of L’Aquila, 67100 L’Aquila, ItalyDental Clinic Center, Kyungpook National University, Daegu 41940, KoreaDefense & Safety ICT Research Department, Electronics and Telecommunications Research Institute, Daejeon 34129, KoreaFault detection and diagnosis is one of the most critical components of preventing accidents and ensuring the system safety of industrial processes. In this paper, we propose an integrated learning approach for jointly achieving fault detection and fault diagnosis of rare events in multivariate time series data. The proposed approach combines an autoencoder to detect a rare fault event and a long short-term memory (LSTM) network to classify different types of faults. The autoencoder is trained with offline normal data, which is then used as the anomaly detection. The predicted faulty data, captured by autoencoder, are put into the LSTM network to identify the types of faults. It basically combines the strong low-dimensional nonlinear representations of the autoencoder for the rare event detection and the strong time series learning ability of LSTM for the fault diagnosis. The proposed approach is compared with a deep convolutional neural network approach for fault detection and identification on the Tennessee Eastman process. Experimental results show that the combined approach accurately detects deviations from normal behaviour and identifies the types of faults within the useful time.https://www.mdpi.com/1424-8220/19/21/4612autoencoderlong short-term memoryrare eventfault detectionfault diagnosistime delay |
| spellingShingle | Pangun Park Piergiuseppe Di Marco Hyejeon Shin Junseong Bang Fault Detection and Diagnosis Using Combined Autoencoder and Long Short-Term Memory Network Sensors autoencoder long short-term memory rare event fault detection fault diagnosis time delay |
| title | Fault Detection and Diagnosis Using Combined Autoencoder and Long Short-Term Memory Network |
| title_full | Fault Detection and Diagnosis Using Combined Autoencoder and Long Short-Term Memory Network |
| title_fullStr | Fault Detection and Diagnosis Using Combined Autoencoder and Long Short-Term Memory Network |
| title_full_unstemmed | Fault Detection and Diagnosis Using Combined Autoencoder and Long Short-Term Memory Network |
| title_short | Fault Detection and Diagnosis Using Combined Autoencoder and Long Short-Term Memory Network |
| title_sort | fault detection and diagnosis using combined autoencoder and long short term memory network |
| topic | autoencoder long short-term memory rare event fault detection fault diagnosis time delay |
| url | https://www.mdpi.com/1424-8220/19/21/4612 |
| work_keys_str_mv | AT pangunpark faultdetectionanddiagnosisusingcombinedautoencoderandlongshorttermmemorynetwork AT piergiuseppedimarco faultdetectionanddiagnosisusingcombinedautoencoderandlongshorttermmemorynetwork AT hyejeonshin faultdetectionanddiagnosisusingcombinedautoencoderandlongshorttermmemorynetwork AT junseongbang faultdetectionanddiagnosisusingcombinedautoencoderandlongshorttermmemorynetwork |