Apeliotes - UAV flight control system
-=[BEST VIEWED IN MOZILLA]=-
WE HAVE MOVED TO...
Please update your links and bookmarks.
Apeliotes
is an autonomous magnetometer based UAV flight control system using a high-performance low-power ARM7 cpu core. It uses a
three axis magnetometer to measure the Earth's magnetic field and develop an AHRS solution. The system
uses a more integrated design than the traditional IMU, AHRS, GPS etc
into controller design. The design enables the processor to
have access to each of the sensors information to produce accurate
solutions.
Apeliotes is being developed in
The development began at the start of 2005 as part of my Dissertation
year for an honors degree as a Bachelor of Applied Science
(BAppSc). The research will continue toward a MSc during 2006 by
increasing the capabilities of the flight controller.
Code, schematics and board files will be available open source
The projects homepage at the University of Otago is:
http://www2.physics.otago.ac.nz/research/electronics/uav/
THE LATEST: (site last edited 22/11/06)
We've upgraded our hosting in preparation of releasing some of our work. Please now use the new website at http://www.uavdev.com
Please update your bookmarks and links to the new address. Some of the advantages we have gained with the new website include: phpbb forum, visitor editable event calendar (submit your upcomming event), users can post comments on events and upload attachments for the event (upload your photos from the event afterwards), don't be shy we now have 1.5TB of monthly bandwidth and 100GB of storage. We also have live chat with other visitors to the site, links, project info and downloads, and more still to come. I hope you enjoy the new site. Heres a screenie of what to expect.
Hardware in the loop attitude testing in simulator. (22-Nov-2006)
The simulator is calculating body frame magnetic and gravitational vectors based on the attitude of the aircraft in the simulator. These vectors are passed to the control board as if they where gps measurements, except they occur at 20Hz (or what ever is desired). The image below shows the onboard calculated attitude vs the simulator calculated actual attitude.
Made a physics simulator with GUI . (24-Oct-2006)
Ok we've made a physics simulator so that we can test and analyse new control systems out before we fly. Theres not a lot of good flying days down here so it's been essential. It's also useful for learning the dynamics of an aircraft to make a control system. Here's a screenshot of the GUI.
Better development cabling. (18-Aug-2006)
The plane can now be programmed and debugged/talked-to using USB, with a much neater cableing design. It is also now USB bus powered, which is much easier. The red light indicating when power is sourced from the USB line.
New board design testing well. (28-Mar-2006)
The new board is proving to be more reliable. It's also smaller and lighter, with much more functionality. Here's a few pictures showing that it is possible to etch the new design:
The origonal system prooved to be unreliable once
the plane flew out of range, There was a small probability of the plane
switching back into remotely piloted mode. So the origonal electronics
(johnson counter, octal buffers and D-flip-flops) have been replaced
with one small CPLD. This also means that a servo extension board can
easily be accomodated run off one of the servo outputs, enabling the
use of any number of PWM outputs.
GPS is up and running with logging. (10-Dec-2005)
The board has been modified so that it now includes a SD memory card
slot so that all telemetry information can be easily stored and later
recovered.
Heres a link to download the video footage: (22MB)
(right-click the link > and select "save as" if you don't want it streaming)
bravenet.com