On Thu, 10 Jul 2008 23:18:27 -0500 "Curtis Olson" <[EMAIL PROTECTED]> wrote:
> I just posted a 60+ Mb movie to my web page. As with most of my movie > posts, it might not be completely worth the time to download. :-) > > http://baron.flightgear.org/~curt/tmp/20080710.AVI > > This isn't directly FlightGear related, but it does tie in a couple ways. > The flying wing UAS is running several code modules straight out of the > FlightGear/SimGear project ... (1) our xml parser (2) our property system > (3) the FlightGear autopilot system (4) an adaption of the FlightGear route > following system. This code has been ported to a small gumstix embedded > computer which runs Linux, and the *gear code is pretty much verbatim except > for a few small tweaks to remove unneeded dependencies. In order to > highlight it's heritage, the code running on the UAS is called "MicroGear". > It is very nice to be able to load and parse xml configurations files. The > property system is very convenient for exchanging data between modules and > for referencing configuration data parsed out of the xml files. And the > coolest thing (I think) is that the UAS is running a direct port of the > FlightGear autopilot system ... and it works quite well in the real world. > In fact, if you have a decent flightgear model of your aircraft you can do a > lot of gain tuning in the simulator and then just copy your autopilot.xml > file over to the UAS and it just might work out of the box ... I'm 1 for 2 > on that. > > The display in the movie is developed by John Wojnarowski (of LFS > Technologies and a contributor to the FlightGear project -- > http://www.lfstech.com) You can probably figure this out, but there is a > live com link between the UAS and the ground station. The UAS is > continually blasting flight data, autopilot data, and other status and > health data down to the ground station, and the ground station can reply > with commands (i.e. to change the route, fly at a different altitude, come > home, etc.) The com link is a wireless serial connection (900mhz) so it > doesn't have enough bandwidth to send all the data at full rate. Thus you > can see the live display is not perfectly smooth. But the data is captured > at full rates onboard, so all this data can be replayed nice and smoothly > later, back at the shop. > > I don't have time to write a book about the LFS glass display tonight, but > it's been a tremendously useful tool for tuning the autopilot, monitoring > live flights, and analyzing flight performance. For instance, the altitude > hold module is comprised of three stages ... Stage 1 outputs a target rate > of climb based on altitude error. Stage 2 outputs a target pitch angle > based on the rate of climb error. Stage 3 outputs an elevator deflection > based on the pitch angle error. So in the LFS glass display, you can see > current altitude, target altitude (drawn as a bug), current rate of climb, > target rate of climb (drawn as a bug), the PFD shows your pitch angle, and > an area on the right side shows all the control surface deflections. The > coolest part (I think) is the flight director vbars that show target pitch > angle (and also target roll.) So if the autopilot is doing it's job, the > yellow "bird" should sit right inside the green "vbars". It's fun to watch > all the different components interact. The bird is always chasing the vbars > at least a little bit ... especially on turbulent days. So being able to > see a nice real time graphical representation of the key inputs and outputs > for each stage of the autopilot is tremendously helpful for isolating which > stage might still need some tuning and with a little experience you can know > which parameter to tune and in what direction. > > This has been a fun project ... a few cool toys have landed on my desk to > play with, this spring I ended up on a NOAA research ship 1000nm from the > nearest point of land to do test flights of this UAS. And it's really neat > to see how code that has been developed and tuned and refined in a pure > simulation environment can be moved over to an embedded computer on a real > aircraft and it works just as well over there too. > > Curt. > -- > Curtis Olson: http://baron.flightgear.org/~curt/ > The display is really interesting , looks like the Garmin G1000 Ive been working on , http://www3.telus.net/sydadams/ (stalled due to difficulty with GPS moving map). How do we convince John to get this into Flightgear ? Cheers -- Syd&Sandy <[EMAIL PROTECTED]> ------------------------------------------------------------------------- Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! Studies have shown that voting for your favorite open source project, along with a healthy diet, reduces your potential for chronic lameness and boredom. Vote Now at http://www.sourceforge.net/community/cca08 _______________________________________________ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
