Re: [Paparazzi-devel] Estimator and Autopilot Handover Questions

[Pascal Brisset]

Hi David,

there is no "running" estimator of the attitude using infrared. 
Measurements are directly used, without any filtering. The only low pass 
filter is the one directly done on the output of the ADC (average over 
16 last samples). So the "convergence" time is only due to the sensors 
themselves

--Pascal

David Bryant wrote:
> Hi,
>
> I have a concern in regard to our novel use of paparazzi that I am 
> trying to address.
>
> On launch, our aircraft sits in a cannon mount that prevents the 
> thermopiles from having a view of the surroundings. Paparazzi is 
> instructed to keep the control surfaces fixed at neutral. Then the 
> cannon launches the aircraft, and we detect this launch using an 
> accelerometer. Some small period later we tell the autopilot to take 
> control of the plane.
>
> My concern is whether there will be a window of time between:
>
> The thermopiles acquiring a clear view of the surroundings and 
> consequently stabilising the estimator.
>
>  AND
>
> The duration that the aircraft can be self-stable relying on fixed 
> control surfaces.
>
> Some questions:
>
> When is the estimator running in paparazzi? Always?
>
> We know the initial attitude of the aircraft; is it possible to 
> (re)initialise the estimator using these values just prior to letting 
> the autopilot control the aircraft?
>
> Approximately what is the convergence time of the estimator from 
> thermopile input?
>
> Thanks,
> Dave
>
>
> _______________________________________________
> Paparazzi-devel mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/paparazzi-devel