Key points are not available for this paper at this time.
The copious volumes of data collected incessantly from diverse systems present challenges in interpreting the data to predict system failures. The majority of large organizations employ highly trained experts who specialize in using advanced maintenance equipment and have specific certification in predictive maintenance (PdM). Prognostics and health management (PHM) connects research on deterioration models to strategies for PdM. Prognostics refer to the process of estimating the time until failure and the associated risk for one or more current and potential failure modes. Prognostics aim to provide guidance by alerting to imminent failures and predicting the remaining useful life (RUL). This eventually leads to improved availability, dependability, and safety, while also reducing maintenance costs. This research work diverges from existing established prognostic methodologies by emphasising the use of hybrid prognostics to predict the future performance of an aircraft system, especially the point in which the system will cease to operate as intended, often referred to as its time to failure. We have developed a new method that combines a physics-based model with the physics of failure (PoF) and a multiple-layered hyper-tangent-infused data-driven approach. The results are useful. The authors retrieved datasets for analysis using a laboratory aircraft fuel system and simulation model. Consequently, the comparative results demonstrate that the proposed hybrid prognostic approach properly estimates the RUL and demonstrates strong application, availability, and precision.
Felixfu et al. (Thu,) studied this question.