 |
International Society of Science and Applied Technologies |
| Predictive Maintenance Decisions Based on System Remaining Useful Life Prediction and Three Inspection Strategies | ||||
| Author | Xufeng Zhao
|
|||
| Co-Author(s) | Lubing Wang; Mingchih Chen; Yu Yang
|
|||
| Abstract | Remaining useful life (RUL) prediction and predic- tive maintenance decisions are two significant research problems in system prognostics and health management (PHM). However, most existing studies have executed these two problems sepa- rately and hierarchically, and rarely integrate the two. To solve this problem, this paper proposes a complete dynamic predictive maintenance decisions framework based on system RUL pre- diction and three inspection strategies to ensure the safety and reliability of the system. Considering RUL prediction aspect, a hybrid deep learning model is proposed to effectively predict system RUL, which combines a one-dimensional convolutional neural network (CNN) and a bidirectional long short-term memory network with an attention mechanism (Bi-LSTM-AM). Meanwhile, a Bayesian optimization method is proposed to further improve RUL prediction performance. Regarding the post-prognostic aspect, we consider periodic, descending, and state-based inspection strategies to help decision makers for- mulate predictive maintenance decisions. Finally, experimental results demonstrate the superior performance of the proposed dynamic predictive maintenance decision framework.
|
|||
| Keywords | Remaining useful life; Predictive maintenance; Inspection strategies; Deep learning. | |||
| Article #: RQD2024-119 | ||||
Proceedings of 29th ISSAT International Conference on Reliability & Quality in Design |