 |
International Society of Science and Applied Technologies |
| System Reliability Prediction Via Long Short-Term Memory for Manufacturing System | ||||
| Author | Cheng-Hao Huang
|
|||
| Co-Author(s) | Ding-Hsiang Huang; Yi-Kuei Lin
|
|||
| Abstract | With the development of system reliability evaluation, more and more studies are published by considering more attributes or adopting different calculation methods. However, existing deep learning (DL) approaches for predicting system reliability of a stochastic manufacturing network (SMN) only considering single attribute, such as time or machine failure. A more comprehensive consideration about an SMN is deliberated in this study. The Long Short-Term Memory is then utilized to construct the prediction model to process time series of an SMN data. Through the experimental results, the proposed prediction model outperforms than the existing method and other DL method in terms of root mean square error and mean absolute error. Moreover, the parameter of an SMN and hyperparameter of the proposed prediction model are discussed to investigate the optimized combination for system reliability prediction.
|
|||
| Keywords | stochastic manufacturing network (SMN), system reliability, long short-term memory (LSTM) | |||
| Article #: RQD2024-172 | ||||
Proceedings of 29th ISSAT International Conference on Reliability & Quality in Design |