Re: [Paparazzi-devel] Control loops in simulink

[Hector Garcia de Marina]

Hello Niki,

as far as I know, XPlane does not work with coefficients. What do you mean? In the past I used 

the coefficients for tuning the control (at trimmed state) and test the control in XPlane. I realized that XPlane is not intended (at least not the FAA certified version) for real time simulation (software in the loop, hardware in the loop, etc) as it has problems with time-step and other "telemetry" issues (there are many artifacts that ruins your propagation in the equations). So I had to build my own 6DoF simulator to fit with my firmware, matlab, etc.

Of course you can use the "navigation values" (like attitude) from XPlane (like many people do for control proposes), but if you use "raw data" like gyros, accelerations, etc, you will face many problems.

Xflr5 ( http://xflr5.sourceforge.net/xflr5.htm ) is very friendly, and it aims to have more accuracy than AVL / Tornado for low Re number (small/mid UAVs), but at the moment is still in development changing every other day.

About ardupilot, I found it lack of power (compared with STM32 Cortex M3) for complex GNC algorithms, and I want to apply vision algorithms in a near future (Overo can handle it) without changing the HW.

Cheers.

Héctor.

On Wed, Jun 29, 2011 at 9:04 AM, Ing. Niki Regina <address@hidden> wrote:

Hy Jesus and Hector,

 

I follow the same procedure of Hector to find to aerodynamic coefficent and aircraft model. Once you have it you can make you airplane with these coefficents in xplane and start to tune your PID gains.

I don't know xflr5. 

I have never used piccolo autopilot yet. I hope to do it in the future.

 

I am using now ardupilot and you've got in this HW more interfaces than with paparazzi.

 

Niki

----- Original Message -----

From: Hector Garcia de Marina

To: address@hidden

Sent: Wednesday, June 29, 2011 8:13 AM

Subject: Re: [Paparazzi-devel] Control loops in simulink

Hello Jesus,

what kind of process are you using to obtain the aerodynamic coefficients / aircraft model? Right now I am using AVL and xflr5 (really nice for low Re number, but at the moment with many bugs).

About the PID tunning, I guess paparazzi philosophy is kinda different to piccolo from cloudcap. I have used piccolo in the past and I have to say is very friendly and quickly you can accustom to it. As you, I took paparazzi HW as starting point (you will save loads of time and it is powerful enough with Overo, even for light algorithms you do not need it), but simulator, firmware, matlab interfaces, etc, I had to write them on my own to fit my GNC development requirements.

By the way, from what univeristy are you? are you from Spain?

Cheers.

Héctor.

On Wed, Jun 29, 2011 at 12:21 AM, Jesus Martin Sanchez <address@hidden> wrote:

Hello:

  I want to do it as an educational experience, I work in robotic lab at the university and we want to use paparazzi in other projects changing control laws, using fault tolerant control... etc and I think that it could be a good point of item. 

  My PID tuning system, is not automatic. Have you used el CloudCap Piccolo tuning system? It´s a very easy way of tune the autopilot without risk I think it will be very usefull for everybody. I know how do a dinamic model with relative good precision, (i´m sure that better than the AVL and the propulsion model that use Piccolo) but I only ask for the paparazzi loop in simulink, because I would save myself a lot of time. If not, nothing happens, I´m going to need more time, but I´m going to do it.

  Thanks

  Jesús

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

--
Héctor

---

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

LA RICERCA C’È E SI VEDE:
5 per mille all'Università di Bologna - C.F.: 80007010376
http://www.unibo.it/Vademecum5permille.htm

Questa informativa è inserita in automatico dal sistema al fine esclusivo della realizzazione dei fini istituzionali dell’ente.

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

--
Héctor