Improving GPS accuracy

Comments

• Bean Posts: 8,129

  2006-10-13 17:15 edited 2006-10-13 17:15

  I don't·think there is any easy way to do that.
  You're asking an awful lot from an $80 GPS. I'm surprised it's that accurate.
  
  Bean.
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  Cheap used 4-digit LED display with driver IC·www.hc4led.com
  
  Low power SD Data Logger www.sddatalogger.com
  SX-Video Display Modules www.sxvm.com
  
  Don't mistake experience for intelligence. And vis-vera.
  
  
  Post Edited (Bean (Hitt Consulting)) : 10/13/2006 6:39:29 PM GMT
• RDL2004 Posts: 2,554

  2006-10-13 17:47 edited 2006-10-13 17:47

  I have to agree with Bean. I have an eTrex Legend C and·I don't think its basic GPS accuracy is much better (maybe when WAAS is available), and it cost over $250.
  
  
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  - Rick
• The Captain Posts: 60

  2006-10-13 18:48 edited 2006-10-13 18:48

  Usless you have access to US military GPS equipment, some with accuracies to within +/- 1in or less, your beat, sorry.
• Paul Baker Posts: 6,351

  2006-10-13 19:34 edited 2006-10-13 19:34

  Actually there are a couple methods to provide better accuracy, both require puchasing a new receiver. DGPS and WAAS enabled receivers obtain·an additional source of information to provide better error correction. DGPS uses ground based "beacons" which have a fixed known location, the error it receives between its reported location and known location is transmitted and picked up by DGPS receivers. WAAS is the newer version and uses a geostationary sattelite, the ground based beacons compute thier errors, report it to the WAAS satellite which the WAAS enabled GPS reciever uses to calculate the local error from finding the data of the nearest ground beacon.
  
  BTW you aren't getting 5 m resolution, a GPS can only estimate its error. 10m is pretty much the best that can be obtained from a standard GPS.
  
  Here is a table of the accuracies availible from the different systems:

  GPS     10m
  DGPS    1-5m (depending on how close you are to a beacon)
  WAAS    3m
  LAAS    1.5m
  

  
  LAAS is like WAAS, but you must be closer to the beacon in order to use it, also it only used near airports to provide airplanes with accurate landing information.
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  Paul Baker
  Propeller Applications Engineer
  
  Parallax, Inc.
  
  Post Edited (Paul Baker (Parallax)) : 10/13/2006 8:02:47 PM GMT
• Paul Baker Posts: 6,351

  2006-10-13 20:05 edited 2006-10-13 20:05

  Addendum, the Parallax GPS is a WAAS receiver. This is the best you can hope to achieve in accuracy (DGPS can be more accurate, however the regions where this higher accuracy is obtainable are quite small).
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  Paul Baker
  Propeller Applications Engineer
  
  Parallax, Inc.
• Fe2o3Fish Posts: 170

  2006-10-13 20:36 edited 2006-10-13 20:36

  WAAS availability will get you only +/- 3-meters @ 95% of the time.
  US Military GPS equipment, while much more accurate, does require
  the ocassional decryption key to get this accuracy.
  
  If you need high accuracy of certain points then averaging of lat/lon
  readings for a period of time will get you the accuracy. If you need
  this accuracy while moving then I don't see how you'll do it with the
  Parallax GPS module.
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  -Rusty-
  --
  Rusty Haddock = KD4WLZ = rusty@fe2o3.lonestar.org
  **Out yonder in the Van Alstyne (TX) Metropolitan Area**
  Microsoft is to software what McDonalds is to gourmet cooking
• Washer Medic Posts: 39

  2006-10-14 03:44 edited 2006-10-14 03:44

  How feasible would it be to set up a small scale beacon using something like the 433 transmitters at each corner?
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  There's nothing a good wack with a hammer won't fix
  
  Darn I let the white smoke out again
• AIman Posts: 531

  2006-10-16 14:00 edited 2006-10-16 14:00

  It wouldn't work. This is a competition where the robot must navigate over given waypoints on a course. The course data·isn't availble until 5 minutes prior to start of the competition. At this point the location hasn't been released and rules forbid using, scanning or navigating the location once released.
  
  I did find out some more things about GPS and some ways to increase accuracy which I thought people would want to know. By using these corrections factors in the code the accuracy is greatly improved.
  
  First,·use microseconds for timing. The smaller the timing units used the better. So by calculation out the next GPS signal at one second based on microseconds you get a more accurate fix on your location.
  
  Second, use division out to 1,000 decimal points or more for all math calculations. Again, the more specific the better.
  
  Third, Adjust the altimeter to the barometeric pressure and location for that day. This data is available from aviation resources and is transmitted 24 hours a day via radio for aviation uses.·For example an airport might be at 392.7 feet above sea level with a pressure of 33.4 for the day.·This data will reflect on the GPS accuracy and can be easily attained. These adjustments will also be a signifcant increase in accuracy.
  
  Finally, run a projection of your location based on the last known cordinates. Example, if the GPS shows you moving at 30 mph then calculate out, based on microseconds, how far you should travel at 30 mph based on·inches (or cm). Figure out from that what should be your new location·and then use that to help eliminate errors by cross checking your GPS data - new and old.
• John R. Posts: 1,376

  2006-10-16 21:06 edited 2006-10-16 21:06

  In terms of using "dead reckoning" (time * speed = distance) you may also want to consider using accelerometers to track your position.
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  John R.
  
  8 + 8 = 10
• Paul Baker Posts: 6,351

  2006-10-16 23:23 edited 2006-10-16 23:23

  Are you provided the precise coordiates of the starting position? If so you can use it and the measured reading to provide some additional accuracy (that may drift after a while if the race takes a long time).
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  Paul Baker
  Propeller Applications Engineer
  
  Parallax, Inc.
• Kevin Wood Posts: 1,266

  2006-10-17 01:20 edited 2006-10-17 01:20

  You should also verify that your gps is referencing the correct datum for the area you are in.
• AIman Posts: 531

  2006-10-17 02:58 edited 2006-10-17 02:58

  I don't know if the starting point is on the waypoint list or not. However that will be easy enough to obtain during the 5 minutes to race start.
• Paul Baker Posts: 6,351

  2006-10-17 05:50 edited 2006-10-17 05:50

  Well if you do obtain the "true" location of the starting point (Not GPS generated, but the actual coordinate), you can compare that with what your GPS provides to get an offset. This will act in the same way a beacon does, and with the other techniques you mentioned should get you going.
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  Paul Baker
  Propeller Applications Engineer
  
  Parallax, Inc.
• Fe2o3Fish Posts: 170

  2006-10-17 06:46 edited 2006-10-17 06:46

  Alas, that offset is only good for a short period of time -- the time it takes for
  the list of tracked GPS satellites to change. Once even a slightly different set
  of satellites is in the calculations (even by one) the offset will be, usually, quite
  different. I believe the time a given offset is valid is somewhere around 12-15 minutes.
  If this offset were good for the rest of time, WAAS and DGPS would be essentially
  useless.
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  -Rusty-
  --
  Rusty Haddock = KD4WLZ = rusty@fe2o3.lonestar.org
  **Out yonder in the Van Alstyne (TX) Metropolitan Area**
  Microsoft is to software what McDonalds is to gourmet cooking
• AIman Posts: 531

  2006-10-18 02:06 edited 2006-10-18 02:06

  I agree that it is not a permanet solution, however every little bit helps. If for any reason motion is stopped for a few seconds then everything can recalibrate.
  
  Not perfect but helpful.