Dynamic Data Driven Methods for Self-Aware Aerospace Vechicles
Abstract
This project has developed mathematical and computational foundations of Dynamic Data Driven Application Systems (DDDAS) methods that combine physics-based and data-driven perspectives. The research has been motivated by and demonstrated in the particular context of structural health management for a self-aware unmanned aerial vehicle (UAV); however, the developed methods and approaches are broadly applicable across DDDAS applications. Significant outcomes of the project include a new methodology for creating a Predictive Digital Twin, using component-based reduced modeling and interpretable machine learning. The Digital Twin is built from a library of component-based reduced-order models that are derived from high-fidelity finite element simulations of the vehicle in a range of pristine and damaged states. In contrast with traditional monolithic techniques for model reduction, the component-based approach scales efficiently to large complex systems, and provides a flexible and expressive framework for rapid model adaptation both critical features in the digital twin context. The project demonstrated a Digital Twin use case for rapid structural health assessment and dynamic mission re-planning. Another significant outcome is a new suite of approaches for managing sensors and sensing strategies, including detection and correction of sensor errors, multiple information source fusion, and optimization of sensor locations to support real-time operational decision. Finally, the project achieved design and construction of aight test vehicle that serves as a DDDAS testbed.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 15, 2020
- Accession Number
- AD1110455
Entities
People
- David Kordonowy
- Douglas Allaire
- Karen Willcox
Organizations
- Massachusetts Institute of Technology