There's an algorithm for everything, it seems
Like eagles in flight, sailplanes depend on finding thermals, or pockets of rising air, to keep them aloft. Finding those regions of lift comes naturally for birds, but low-power autonomous sailplanes need a way to quickly find the conditions that will keep themselves aloft.
The crux of autonomous soaring is finding regions of lift, said Dan Edwards, aerospace engineer with the Naval Research Lab. So NRL is working with researchers from Penn State to test an algorithm for cooperative and autonomous soaring of unmanned sailplanes. With several communicating sailplanes, the chances of quickly finding this lift increase, and all the vehicles can stay airborne longer.
The Autonomous Locator of Thermals algorithm uses technologies tested and developed by both Penn State and NRL to share vehicle data -- such as sailplane location, longitude, latitude, altitude -- with the rest of the flock, Edwards said. Sailplanes within the flock can then move autonomously to a location where one sailplane has found sufficient lift.
The project “combines data from multiple autonomous soaring aircraft to make a more complete measurement of the local atmospheric conditions,” said Edwards. “This atmospheric map is then integrated to guide both aircraft toward strong lift activity quicker than if it was just a single aircraft -- a technique very similar to that used by a flock of soaring birds.”
Using the algorithm to share data on the location of thermals, the sailplanes were able to fly for hours despite having onboard batteries that provide only enough energy for a few minutes of powered flight.
While the demonstration tested just two sailplanes together, the next step is to test four, Edwards said. At the moment, there is no technical barrier to testing 100 devices together -- the practical limits are resources and manpower.
Posted by Mark Pomerleau on Jan 28, 2016 at 10:36 AM