A Hybrid LSSVR/HMM-Based Prognostic Approach
In a health management system, prognostics, which is an engineering discipline that predicts a system’s future health, is an important aspect yet there is currently limited research in this field. In this paper, a hybrid approach for prognostics is proposed. The approach combines the least squares s...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2013-04-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | http://www.mdpi.com/1424-8220/13/5/5542 |
| _version_ | 1818039672663703552 |
|---|---|
| author | Xianhui Liu Chundi Mu Qing Li Zhijuan Liu |
| author_facet | Xianhui Liu Chundi Mu Qing Li Zhijuan Liu |
| author_sort | Xianhui Liu |
| collection | DOAJ |
| description | In a health management system, prognostics, which is an engineering discipline that predicts a system’s future health, is an important aspect yet there is currently limited research in this field. In this paper, a hybrid approach for prognostics is proposed. The approach combines the least squares support vector regression (LSSVR) with the hidden Markov model (HMM). Features extracted from sensor signals are used to train HMMs, which represent different health levels. A LSSVR algorithm is used to predict the feature trends. The LSSVR training and prediction algorithms are modified by adding new data and deleting old data and the probabilities of the predicted features for each HMM are calculated based on forward or backward algorithms. Based on these probabilities, one can determine a system’s future health state and estimate the remaining useful life (RUL). To evaluate the proposed approach, a test was carried out using bearing vibration signals. Simulation results show that the LSSVR/HMM approach can forecast faults long before they occur and can predict the RUL. Therefore, the LSSVR/HMM approach is very promising in the field of prognostics. |
| first_indexed | 2024-12-10T08:02:22Z |
| format | Article |
| id | doaj.art-d4a0f2c18ba540149cf070aa167465e5 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-12-10T08:02:22Z |
| publishDate | 2013-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-d4a0f2c18ba540149cf070aa167465e52022-12-22T01:56:45ZengMDPI AGSensors1424-82202013-04-011355542556010.3390/s130505542A Hybrid LSSVR/HMM-Based Prognostic ApproachXianhui LiuChundi MuQing LiZhijuan LiuIn a health management system, prognostics, which is an engineering discipline that predicts a system’s future health, is an important aspect yet there is currently limited research in this field. In this paper, a hybrid approach for prognostics is proposed. The approach combines the least squares support vector regression (LSSVR) with the hidden Markov model (HMM). Features extracted from sensor signals are used to train HMMs, which represent different health levels. A LSSVR algorithm is used to predict the feature trends. The LSSVR training and prediction algorithms are modified by adding new data and deleting old data and the probabilities of the predicted features for each HMM are calculated based on forward or backward algorithms. Based on these probabilities, one can determine a system’s future health state and estimate the remaining useful life (RUL). To evaluate the proposed approach, a test was carried out using bearing vibration signals. Simulation results show that the LSSVR/HMM approach can forecast faults long before they occur and can predict the RUL. Therefore, the LSSVR/HMM approach is very promising in the field of prognostics.http://www.mdpi.com/1424-8220/13/5/5542prognosticsleast squares support vector regressionhidden Markov modelremaining useful life |
| spellingShingle | Xianhui Liu Chundi Mu Qing Li Zhijuan Liu A Hybrid LSSVR/HMM-Based Prognostic Approach Sensors prognostics least squares support vector regression hidden Markov model remaining useful life |
| title | A Hybrid LSSVR/HMM-Based Prognostic Approach |
| title_full | A Hybrid LSSVR/HMM-Based Prognostic Approach |
| title_fullStr | A Hybrid LSSVR/HMM-Based Prognostic Approach |
| title_full_unstemmed | A Hybrid LSSVR/HMM-Based Prognostic Approach |
| title_short | A Hybrid LSSVR/HMM-Based Prognostic Approach |
| title_sort | hybrid lssvr hmm based prognostic approach |
| topic | prognostics least squares support vector regression hidden Markov model remaining useful life |
| url | http://www.mdpi.com/1424-8220/13/5/5542 |
| work_keys_str_mv | AT xianhuiliu ahybridlssvrhmmbasedprognosticapproach AT chundimu ahybridlssvrhmmbasedprognosticapproach AT qingli ahybridlssvrhmmbasedprognosticapproach AT zhijuanliu ahybridlssvrhmmbasedprognosticapproach AT xianhuiliu hybridlssvrhmmbasedprognosticapproach AT chundimu hybridlssvrhmmbasedprognosticapproach AT qingli hybridlssvrhmmbasedprognosticapproach AT zhijuanliu hybridlssvrhmmbasedprognosticapproach |