Accuracy and precision of the Parallax GPS module
SteveWoodrough
Posts: 190
The text of the Parallax GPS (PGPS) spec sheet states:
On average, the Module has a +/-5 meter position accuracy and a +/-0.1 meter per second velocity accuracy......
Primarily due to satellite geometry, measuring altitude using GPS may introduce an accuracy error of 1.5 times the receiver’s position accuracy (in the case of our GPS Receiver Module, this corresponds to about +/-20 meters in the vertical direction).
Assuming I have a good signal does this mean the receiver needs to move +/-5 meters to sense a new position or is the accuracy more on an absolute scale?
For example:·· The tree in my back yard is at a specific·Long and Lat coordinate.· Should I assume that the PGPS might differ from another GPS (say my Garmin Nuvi) by 5 meters?· That being the case will the·PGPS be precise and repeatable?· Said another way:· If I record the PGPS·location of the tree given by the PGPS then move the PGPS 100 meters and then return the PGPS back to the same tree, will the PGPS record the same position of the tree?·· If I then turn the unit off, and come out the next day will I get the same position of the tree from the previous day?· All of this assumes that the tree is small and does not otherwise interfere with the GPS signal.
Thank You,
Steve
On average, the Module has a +/-5 meter position accuracy and a +/-0.1 meter per second velocity accuracy......
Primarily due to satellite geometry, measuring altitude using GPS may introduce an accuracy error of 1.5 times the receiver’s position accuracy (in the case of our GPS Receiver Module, this corresponds to about +/-20 meters in the vertical direction).
Assuming I have a good signal does this mean the receiver needs to move +/-5 meters to sense a new position or is the accuracy more on an absolute scale?
For example:·· The tree in my back yard is at a specific·Long and Lat coordinate.· Should I assume that the PGPS might differ from another GPS (say my Garmin Nuvi) by 5 meters?· That being the case will the·PGPS be precise and repeatable?· Said another way:· If I record the PGPS·location of the tree given by the PGPS then move the PGPS 100 meters and then return the PGPS back to the same tree, will the PGPS record the same position of the tree?·· If I then turn the unit off, and come out the next day will I get the same position of the tree from the previous day?· All of this assumes that the tree is small and does not otherwise interfere with the GPS signal.
Thank You,
Steve
Comments
I'm going to say that it the accuracy from a plus/minus perspective has no consistency. I got readings all around the place on my PGPS. At times I even got a reading totally out of the area. But you will get most readings pretty close and accurate. You'll read various reasons for the accuracy differences. Probably the best use we see today of GPS is with the units people have in cars, boats, etc. Most of these incorporate a mapping function that likely has error correction built in. For instance, it assumes you wouldn't be driving your car through rivers and off of highways. So they attempt to put you on a road "near" where the GPS says you are. A true mapping of the actual point the GPS is indicated would likely get you a ticket for reckless driving. One project I had going (that I back-burnered for a bit), was a cycling route logger. Data collected would be downloaded and fed up to a website that maps your route from collected GPS points. I was going to build my own "error correcting" code that would toss any reading that was ridiculously different from the previous reading. For instance, my bike wouldn't be able to travel 1 mile in 5 seconds, so any reading that was indicating something unusual would get dropped and I would cycle back a re-poll the GPS device. I was doing this anyway when I was not getting complete readings. That is to say I would poll every 5 or 10 seconds (ok for bike speeds), and if a bad reading was taken, it would immediately re-poll. As long as the reading was valid, I would cycle back in 5 to 10 seconds. This may answer your question as to whether you would get the same reading under the same circumstances. I'm saying no.
Rick
No. You will see the position change more gradually, and it will be within 5m of your true position at least 50% of the time..
Should I assume that the PGPS might differ from another GPS (say my Garmin Nuvi) by 5 meters?
If both GPSes are receiving data from the same set of satellites at the same time, the reported positions should be the same.
If I record the PGPS location of the tree given by the PGPS then move the PGPS 100 meters and then return the PGPS back to the same tree, will the PGPS record the same position of the tree?
That depends on how fast you can move. If it takes you an hour to go that far, probably not.
If I then turn the unit off, and come out the next day will I get the same position of the tree from the previous day?
Almost certainly not.
These differences have nothing to do with any particular GPS unit but with propagation anomalies from an an ever-changing array of received satellites. If you take any GPS unit, leave it in one spot, and record it's stated position over the course of a day, you will see that it appears to move. But, if the GPS states a position accuracy of +/- 5m, and if you have a clear view of the sky (very important), then at least half the readings should be contained within a circle of radius 5m, 95% within a 10m radius, and 98.5% within a 12.75m radius (rf: "How GPS Works").
-Phil