We were running some tests on RTK / DGPS / GPS these two weeks. Thought it might be interesting to share our results in image format, although the accuracy numbers are already well known.

Test setup is as below:

Below is RTK result from 3 different days; black:20150716, green:20150721, white:20150723

As RTK accuracy is centimeter level, they overlap pretty well at this zoomed-out view.

White line RTK result is used as frame of reference to compare DGPS and GPS results

Below is DGPS result from 2 different days; red:20150716, cyan:20150723. Result is quite good since distance from DGPS base station is only 1.5Km.

Below is GPS result from 2 different days; red:20150716, cyan:20150723.

The shifted red track is due to ionosphere error that changes over time, which accounts for major error in a typical 2.5m CEP accuracy GPS receiver. It just happened that ionosphere error was less on 20150723, resulting in the cyan track closer to the correct white track.

What’s noteworthy is that the two rounds made on each day overlap quite well, meaning that the GPS receiver is fairly precise, just that the accuracy is still 2.5m CEP.

Below is 20150723 result with another popular GPS receiver brand for comparison. 

Overlapping all previous plots, below figure results; giving a glimpse of potential GPS error on day to day basis.

Views: 4728

Comment by Nikola Rabchevsky on July 26, 2015 at 7:50am

I'd be interested to see how the non-RTK results shift from one week to another or when solar activity is high.

Comment by Doug Walmsley on July 27, 2015 at 8:10am

Do you plan to perform additional tests comparing double and triple band GPS units?  I also feel you results are spread out over a period of 7 days (7/16, 7/21, 7/23) may produce different results than if you had performed all tests simultaneously.  Would be interesting to see how RTK vs Non-TRK (both double and triple band) units tested at the same date and time frames.  You mentioned "another popular GPS brand", can you be more specific in which brand you used for your tests.  Was it a NEO-7M, LEA-6H, NEO-M8N, etc? 

I like your initial report of your tests.  Like to see how each of these common GPS units compare to the RTK versions out there on any given Solar day measuring results of each based on K-Index values.  Keep up the great work. You findings will help hobbyists decide which units will fit there needs based on accuracy and cost (eventually).

Comment by Oliver Huang on July 27, 2015 at 9:58am

Consumer or industrial grade GPS receivers typically have 2.5m CEP accuracy spec. If the same experiment is repeated over 100 day period or more, it’s expect to see the GPS results shifted around the true track bounded by the same 2.5m CEP distribution. 

We don’t have proper double and triple frequency equipment to conduct the suggested testing. We happened to be developing low-cost single-frequency RTK receiver, these results came from some field tests we did. Thought the images might be of interest giving users an idea what GPS error would be like, since most users have no absolute reference and tend to think less scattered result for stationary testing is more accurate, which isn't necessarily true (precision vs accuracy issue).

Not to worry about the other brand receiver. In terms of accuracy, small variation of different receiver behavior riding on top of larger common ionosphere error doesn’t make one receiver better than the other.

Comment by Fahad Jahangir on January 27, 2017 at 11:20am

 Did you collect these line continuously from one end to another as a line feature or you just took two end points and interpolated the lines? 

Comment by Oliver Huang on January 27, 2017 at 2:21pm

Any one doing interpolation from end points will get a straight line. Nope, it's walking the rounds collecting receiver result every second.

Comment

You need to be a member of DIY Drones to add comments!

Join DIY Drones

Groups

Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service