Expedition Navigation Software / Filter GPS outliers?

**Posted by Fredrik** 5/24/2018 6:16 am

I'm using Yacht Devices NMEA 2000 Wifi gateway to send data to my tablet. Most of the time it's a perfect solution, however every now and then i recieve a position with a problem in either lat or long sending my either to the equator or the greenwich meridian for a quick second before returning. As it's only for a second it's not to much of a problem, but it messes up my track and chart settings as it zooms out for the new position not covered by my detailed chart. Is there a way to tell expedition to filter out single outlying updates like this? The Greenwich Meridian

**Posted by DuncanR** 5/24/2018 3:24 pm

Fredrik, I feel for you, this has been a problem for me. Generally the way to fix the underlying problem (and it is a problem, since it can skew your SOG, especially if you have it damped or averaged, as I do) is to reduce the baud rate on the GPS to 12800. Of course, if you have AIS hanging off the GPS, you cannot do that as AIS requires 38400 baud.
I have tried to clean up a logfile with this probiem, and the best way I could think of was to find every outlying data row by subtracting each rows longitude from the previous one and, sorting the file by that difference and deleting all the big-difference rows, then resorting the file. I tried a few complicated VLOOKUPS, but could not get any to work as I wanted. My method is feasible on a small logfile, I wouldnt like to try on a big one. Maybe someone else has a better idea.

**Posted by Nick** 5/24/2018 6:55 pm

It is on the list of things to implement.
I wonder if it is something like a bad GPS antenna. How often does it happen?
It would be interesting to see the NMEA data for one of these. The NMEA sentences should have flags to say whether the data is valid or not.

**Posted by Fredrik** 5/24/2018 9:51 pm

I’m using a 10 Hz NMEA 2000 GPS (point-1) selected as a sole GPS source in expedition, I’ve never had the problem on my fixed plotter using the same GPS, so my guess was the WIFI gateway.

**Posted by Fredrik** 7/08/2018 1:06 am

Any news to this problem with "zero" long/ lat? I'm getting a very strange track from my boat ...

**Posted by Nick** 7/08/2018 3:40 am

That suggests a problem with your GPS. It sounds as if it is occasionally sending 0 as the latitude or longitude.
You might find the HDOP and PDOP are very large values.

**Posted by Fredrik** 7/08/2018 12:32 pm

I’m still guessing it’s the Yacht Devices NMEA 2000 to wifi gateway that has a problem, my plotter reading the NMEA 2000 signal directly from the wired network has a perfect track and the point-1 antenna should be a ”precision” gps ... I’ll check with Yacht Devices ... I guess my plotter could be filtering unexpected zeros allready, but I doubt it does - I’ll check HDOP/ PDOP to

**Posted by DuncanR** 7/08/2018 2:47 pm

Fredrik, I have seen that sort of track variation often. Have you tried changing the baud rate on your GPS from 38400 to 12800?

**Posted by Nick** 7/08/2018 3:33 pm

The NMEA baud rate is fixed at 256k.

Setting a value to zero like this would normally indicate a problem with the GPS receiver or (more likely) the antenna or cable to the antenna. This may show in the DOPs.

**Posted by Fredrik** 9/04/2018 8:23 am

I’m super impressed by the rate of new releases - good work! But are there any news to this problem? My track is a nightmare from all the positions of the original track :-( havn’t found any solution on the hardware side, even if that is likely the problem, wouldn’t it be possible with a software fix to avoid reading the position if it is too far off? From what I understand I’m not the only one with this problem.