Journal
MEASUREMENT
Volume 150, Issue -, Pages -Publisher
ELSEVIER SCI LTD
DOI: 10.1016/j.measurement.2019.107097
Keywords
Health indicator; Remaining useful life prediction; Ensemble learning; Genetic algorithm; Weibull proportional hazards model; Support vector regression
Funding
- National Natural Science Foundation of China [61463021, 61963018]
- Young Scientists Object Program of Jiangxi Province in China [20144BCB23037]
- Natural Science Foundation of Jiangxi Province in China [20181BAB202020]
- Science and Technology Project in Education Department of Jiangxi Province in China [GJJ14420, GJJ180494]
- Doctoral Scientific Research Foundation of Jiangxi University of Science and Technology in China [3401223356]
Ask authors/readers for more resources
For mastering device behavior and establishing the mathematical mapping relationship between the degradation process and the operation parameters, we proposed an ensemble RUL prediction model with GA, SVR, and WPHM, and the accuracy and effectiveness of the proposed method were validated by a bearing experiment. To better characterize bearing degradation behavior, a HI construction algorithm was proposed with four metrics of monotonicity, prognosability, trendability and robustness, and RUL prediction was implemented by SVR and WPHM. To verify the superiority of SVR, the performance was compared with NAR-NN, BP-NN, LSTM, GM, ARMA under three criteria including MSE, MAE, and MAPE. Results show that the minimum errors with MSE, MAE, and MAPE appear in SVR, meaning that SVR is the most suitable method in the pseudo operation information prediction. Additionally, the predicted RUL results are basically as same as the actual value by inputting the pseudo operation information into WPHM RUL function. (C) 2019 Elsevier Ltd. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available