A Data-Driven Maintenance Framework Under Imperfect Inspections for Deteriorating Systems Using Multitask Learning-Based Status Prognostics

This paper proposes a data-driven, condition-based maintenance framework (DCBM) for deteriorating equipment under the impact of varying environments and natural aging. The equipment's degradation status is determined by prognostic health monitoring method. Generally, data inspections are imperfect because uncertainties in process, which may prevent reliable evaluation system's deterioration. By utilizing deep learning technique, we construct new stacked autoencoder long short-term memory (SAE-LSTM) network-based multitask model to extract state features from data, then perform multistep forecasting obtain performance failure probability information. developed SAE-LSTM-based achieves prognosis results close actual values, indicates excellent feature extraction capability this model. As result, introduce into optimization process. Probabilistic used as one criteria decisions made with address influence involved prognoses results. effectiveness proposed DCBM illustrated application an engine dataset, more cost-effective than baseline policies.