Design Of Digital Twins for In-Service Support and Maintenance

This research aims to examine the challenges in developing Prognostics and Health Management (PHM) analytics for Digital Twin (DT) use cases in industrial applications, with a particular focus on Multi-Component Degradation (MCD) scenarios. A hybrid methodology, integrating physics-informed and data-driven models, is employed, using limited asset degradation data for model development. Preliminary work includes an analysis of the impact of data quality on Fault Detection and Isolation (FDI) algorithm performance, as well as the proposal of a weighted ensemble hybrid approach for assets experiencing MCD scenarios Preliminary results indicate enhanced diagnostics in asset health management through the use of Physics-Informed models for FDI in MCD scenarios with limited prior degradation data. Expected contributions for this research are the development of physics-informed PHM analytics for DT applications in MCD scenarios, adaptive PHM analytics for evolving asset lifecycles in DT applications, and interpretable DT model analytics for PHM in systems facing Multi-Component Degradation.