Remaining useful life prediction for non-stationary degradation processes with shocks
| Autor: | Youxian Sun, Xiaojie Ke, Zhengguo Xu, Wenhai Wang |
|---|---|
| Rok vydání: | 2017 |
| Předmět: | |
| Zdroj: | Proceedings of the Institution of Mechanical Engineers, Part O: Journal of Risk and Reliability. 231:469-480 |
| ISSN: | 1748-0078 1748-006X |
| DOI: | 10.1177/1748006x17706654 |
| Popis: | Predict the remaining useful life of devices in real time is important to maintenance management. This article addresses a remaining useful life predictive problem where the considered device suffers from non-stationary degradation with shocks. A novel degradation-shock system model subjected to a hidden degradation state is proposed to characterize the degradation process. To estimate the hidden system state, the Kalman filter is modified with optimal estimation. Then, based on Bayes’ theorem, a recursive algorithm is presented to estimate the unknown parameters of the model. With the proposed approach, the hidden state and unknown parameters can be updated at each sampling instant. Subsequently, the analytical solution for the remaining useful life is obtained, while the characteristics of the shocks as well as the uncertainties of the estimated system state are taken into account. Furthermore, the effectiveness of the proposed model is verified by a numerical simulation and a practical case study on a milling machine. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|