Enhancement of Physics-of-Failure Prognostic Models with System Level Features

Abstract

-To truly optimize the deployment of DoD assets, there exists a fundamental need for predictive tools that can reliably estimate the current and reasonably predict the future capacity of complex systems. Prognosis, as in all true predictions, has inherent uncertainty, which has been treated through probabilistic modeling approaches.The novelty in the current prognostic tool development is that predictions are made through the fusion of stochastic physics-of-failure models, relevant system or component level health monitoring data and various inspection results. Regardless of the fidelity of a prognostic model or the quantity and quality of the seeded fault or run-to-failure data, these models should be adaptable based on system health features such as vibration, temperature, and oil analysis. The inherent uncertainties and variability in material capacity and localized environmental conditions, as well as the realization that complex physics-of-failure understanding will always possess some uncertainty, all contribute to the stochastic nature of prognostic modeling. However, accuracy can be improved by creating a prognostic architecture instilled with the ability to account for unexpected damage events, fuse with diagnostic results, and statistically calibrate predictions based on inspection information and real-time system level features. In this paper, the aforementioned process is discussed and implemented first on controlled failures of single spur gear teeth and then on a helical gear contained within a drivetrain system. The stochastic, physics-offailure models developed are validated with transitional run-to-failure data developed at Penn State ARL. Future work involves applying the advanced prognostics process to helicopter gearboxes.

A fault/anomaly system prognosis using a data-driven approach considering uncertainty

“…The uncertainty sources are included in these models and enable to assess the impact of these uncertainties on the remaining-life distribution, in order to make risk-informed decisions. Due to the model complexity, the accumulated degradation is distributed using Monte Carlo simulations, as in [5] and [6]. From the accumulated damage distributions, the remaining life is then predicted with confidence intervals.…”

Condition‐based Maintenance Optimization Using Neural Network‐based Health Condition Prediction

“…The model‐based methods predict health condition using physical models of the components and damage propagation models, such as the bearing prognostics method proposed by Marble et al ,. and the gearbox prognostics methods developed by Kacprzynski et al . and Li and Lee .…”

“…The remaining life prediction uncertainty is required for optimizing CBM activities. The other key challenge is that simulation methods are generally used for the cost evaluation of CBM policies which are based on ANN‐based health condition prediction methods and model‐based prediction methods . They are also used in some CBM methods based on some other data‐driven prediction methods .…”

Estimation of the Remaining Useful Life by using Wavelet Packet Decomposition and HMMs

“…The model-based/reliability fusion approach offers a good accuracy, due to the use of mathematical models, and a long forecasting time horizon by the integration of the experience feedback data. In this field the works of Kacprzynski et al [18,25] deserve to be mentioned. The authors used the Paris law to determine the crack length in a helicopter's gearbox pinion, combined with reliability equations to confirm the degradation phase and to refine the fatigue propagation parameters.…”