A Method for the Exploitation of Microlift for Unmanned Aerial Vehicles
Open Access
- Author:
- Kimmel, Nathan
- Graduate Program:
- Aerospace Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- December 12, 2019
- Committee Members:
- Jacob Willem Langelaan, Thesis Advisor/Co-Advisor
Amy Ruth Pritchett, Program Head/Chair
Mark David Maughmer, Committee Member - Keywords:
- autonomy
soaring
microlift
thermal
thermal wall
thermal stream
thermal strand
autonomous
uav
autonomous soaring - Abstract:
- Disorganized convective structures in the lower part of the atmospheric boundary layer (sometimes called microlift) are recognized as a viable source of energy for manned ultralight sailplanes and high performance hang gliders, and species of soaring birds have been observed to make use of microlift. However, this source of energy remains largely underutilized by autonomous soaring aircraft, and the techniques used to (often incidentally) exploit it are not specifically designed for microlift soaring. Analyzing surveys of the atmosphere and large eddy simulations of the convective boundary layer indicate the presence of long, linear regions of lift that are a potential significant source of energy and present the opportunity for energy extraction without the need to loiter. This thesis investigates techniques for energy extraction from these structures - referred to as spoke-like structures, thermal walls, thermal streams, or thermal stands - that can be implemented onboard an autonomous microlift soaring aircraft. The primary focus of this thesis is a Kalman filter designed to estimate the location, orientation, and characteristics of a thermal strand. This estimator was then integrated into a flight controller and tested in a realistic wind field obtained from a large eddy simulation of the convective boundary layer. The choice of a Kalman filter is supported in this thesis by showing that several basic techniques for following a thermal strand are unstable. Furthermore, an idealized model of a thermal strand is developed for estimator and controller design, and various methods of estimating the initial thermal strand state for the Kalman filter are developed and tested in simulation.