Real-Time Mobility Estimation for Autonomous Rotorcraft to Support Long-Term Autonomy
Author(s)
Sherfield, Sarai N.
Advisor(s)
Editor(s)
Collections
Supplementary to:
Permanent Link
Abstract
This thesis explores real-time mobility estimation for autonomous air vehicles using online machine learning. The kernel of the proposed algorithm is online estimation of the total power required and total power available as a function of airspeed. Applying a Sparse Gaussian Process to estimate the power required at various speeds presents an opportunity to exploit the relationship to assess the utility with the addition of confidence intervals. From this core estimate, all key mobility characteristics can be computed in real-time including uncertainty bounds. Continuous estimation of the performance envelope of an air vehicle is a necessary element for long-term autonomy of air vehicles to ensure survivability.
Sponsor
Date
2024-04-26
Extent
Resource Type
Text
Resource Subtype
Dissertation