Re: [paparazzi-devel] Replies to posts on the digest

[antoine . drouin]

oups... forgot to CC the list

Hi Hugo

Selon Hugo Vincent <address@hidden>:

> The ground station is being written in QT/Free, from www.trolltech.com.
> This allows it to be multi-platform and run on Mac OS X (my operating
> system of choice) as well as linux/unix (or with a license, on M$
> Windows... ).
> I would have built on top of an existing ground station, like the
> paparazzi one, but its inconvenient to use GTK and X11 on Mac OS,
> whereas QT runs natively (without an X server etc, straight on to the
> Mac OS X Aqua engine). I don't particularly like Java GUIs but that was
> the other main possibility.
> I am just learning C++ and QT at the moment, but I hope to have a beta
> suitable for consumption in a month or two. I have got to the point
> where the GUI is laid out (in QT Designer), and most of the 'glue' code
> to run the GUI is working (sortof :), and I am writing the 'back end',
> down links etc.
> Any ideas/suggestions/requests?

I guess the toolkit is not very important. I am currently testing ZINC
(www.openatc.org) with its perl interface - it is based on TK. We are thinking
about writing a binding for ocaml as I don't like perl. I don't like Java too
much neither :)

I have put a screenshot of a PFD made with ZINC at
http://antoine.drouin.free.fr/screenshot.png

You can animate your GUI by replaying our telemetry logs. There is one (old)
example in the CVS. Contact me if you want more.

>
> At some point, I also want to build a link to JSBsim or FlightGear into
> it, for flight visualization when its in the air, and as a hardware in
> the loop simulator, when its on the ground. I am a little worried about
> how well JSBsim will handle simulating such a small plane, an low
> speeds and low reynolds numbers, and also how i will create and
> calibrate a flight model for it. Any ideas?

Well, we'll have to work on this someday. For now we have a dumb fixed wing
flight model.There is a project about fitting the Paparazzi controller in a
quadrirotor airframe and we have student working on
producing a suitable flight model to integrate in our hardware in the loop
simulator.

> > If you like, we can host your document in the paparazzi CVS at
> > savannah.
> OK, that would be great, would it be possible for me to get a login to
> the CVS to that directory (I am not sure if you can do that?) so I can
> change it with out you having to update the CVS manually?

I can add you to the list of developpers if you whish. Open an account at
savannah and tell me your login.

>
> -----------------------------------------------------
> > Message: 3
> > Date: 30 Apr 2004 20:48:12 +0200
> > From: Antoine Drouin <address@hidden>
> > Subject: Re: [Paparazzi-devel] (no subject)
> > To: address@hidden
> > Cc: address@hidden
> > Message-ID: <address@hidden>
> > Content-Type: text/plain
> ...
> > Airspeed:
> > We have build a sensor with a tiny propeller driving a cell phone
> > motor.
> > We amplify and measure the analog voltage produced. The sensor works
> > fine for the small speeds we fly (8-15m/s).
> > We have done nothing yet with these data.
> Hmm, I am sure this works, but it doesn't sound like its the "optimal"
> solution. A pitot tube is much more elegant, more accurate, more
> reliable (no moving parts), is there a reason you haven't tried this.
> You can make one easily, with another pressure sensor attached to a
> tube facing forward at the front of the plane (in 'ram' air). There is
> a simply formula, but i can't remember it (something like V = 1/2 *
> [density of static air (dependent on altitude)] * [ram air pressure
> from pitot tube]^2 ).
>

The pitot works very well for higher speeds. We have tried to sense 10m/s with a
pitot but didn"t succeed in finding a sensitive enough pressure sensor or
good enough amplification.
I had a document on it. I will try to find it again
 
> -----------------------------------------------------
> > Message: 8
> > Date: Sun, 2 May 2004 18:12:08 +0200 (MEST)
> > From: address@hidden
> > Subject: [Paparazzi-devel] conceptual design proposal (without IR
> >         sensors)
> > To: address@hidden
> > Message-ID: <address@hidden>
> > Content-Type: text/plain; charset="us-ascii"
> >
> > [....]
> > As attachement, i sent the planned overall structure of the SW.
> > I plann to use fuzzy logic or a rule system or neuronal logic or
> > genetic algorithm.
> > It's highly possible, that the resulting algorithm use a combination of
> > the previosly mentioned methods.
> > Currently, fuzzy logic seems the simplest and is the preferred for the
> > implementation.
> >
> > Chri
>
> Hi Chris,
>
> This sounds very cool, although, why are you so concerned about radar
> stealth, TEMPEST, radioactivity etc? You make it sound A LOT like a
> military/war UAV! Although I am not really qualified to say, I don't
> think a military UAV should use free GPL'd work -- its use should be
> peaceful. What do the project leaders think about this??
>
> Hugo.
>
>

It's a free project, so everyone can do what he wants with it.
Nevertheless, I really favor peaceful applications and won't take
part in any military oriented development.
Furthemore, I think the military don't need us and already have much more
sophisticated machines.

Cheers

Antoine