paparazzi-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Paparazzi-devel] =?ISO-8859-1?Q? NMEA setup for Papar?= =?ISO-8859-


From: martin
Subject: Re: [Paparazzi-devel] =?ISO-8859-1?Q? NMEA setup for Papar?= =?ISO-8859-1?Q?zzi Autopilot??=
Date: Tue, 25 Mar 2008 23:19:06 +0100 (CET)

Hi David,

> I was hoping to hear Paparazzi Developers were working on NMEA parsers
> or non-proprietary u-blox GPS communication in the code they could
> share. Again, I may have simply missed it so any info appreciated.

It seems that there is the idea that using the ublox protocol (instead
of NMEA) is a "bug". It is not. The ublox is there for a number of
reasons. The protocol itself is well defined and its usage gives a very
good result. The UTM grid that Paparazzi works with internally can be
directly pulled out of the messages without any costly/risky math
conversion. No GPS receiver specific hick-ups in the NMEA protocol
implementation have to be considered.

Also the ublox reveiver itself is there for a reason. The most obvious
is that it gives 4Hz update which is a need for faster-than-Twinstar
aircrafts. Not so obvious is the fact that the dynamic behaviour of the
Kalman filters can be changed - and they have to be. If you do not use
the 4g's aircraft mode, you can perfectly kill a GPS by just flying.
Even in 4g mode you can kill the Kalman filters and get very random
position fixes for some time if you do manual RC aerobatics exceeding
4g. Most important, the ublox is well tested and understood to work.
The GPS receiver is the most important sensor in the aircraft regarding
navigation. A GPS receiver does not give a deterministic result like a
voltage meter. The trajectory you get from a given flight can be very
different depending on the receiver itself. We have tested the ublox
receivers in different dynamic scenarios with a GPS simulator and found
it to be the best fit for our needs.

This does not mean that it has to stay like this. If there is a better
receiver solution, please post it! There seems to be some code to use
NMEA and it would be interesting to see flight test data with other
receivers. Just be aware that changing the GPS receiver is a big change
in the entire system that has to be considered well.

Martin







reply via email to

[Prev in Thread] Current Thread [Next in Thread]