Re: [time-nuts] DGPS@home
On Thu, 24 Nov 2011 13:01:41 +0100 ehydra ehy...@arcor.de wrote: Hi all! I wonder what would be reasonable location accuracy if two cheap same type GPS modules will be several meters apart? I understand that it involves statistical numbers. Any idea? Say for a small robot. On my endless quest to read the whole internet, i stumbled upon this: http://gpspp.sakura.ne.jp/rtklib/rtklib.htm http://gpspp.sakura.ne.jp/rtklib/rtklib_beagleboard.htm I'm not 100% sure, but i think it looks like what you are looking for. Attila Kinali -- Why does it take years to find the answers to the questions one should have asked long ago? ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
Kalman filtering navigation? On Sun, Nov 27, 2011 at 4:53 AM, WB6BNQ wb6...@cox.net wrote: Or an alcohol sensor ! BillWB6BNQ Chris Albertson wrote: snip GPS is never going to be exact. Or I should say you don't know the exact lat. long. for every place you want to go. So to find something like a bear bottle in your refrigerator you need vision Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
Hi Hal - Thanks for your efforts! I just settled down my GPS for the car on my desk (under a brick roof) and left it over night alone. Between evening and morning I wrote 4 locations on paper and later dumped it into Google. So this is a test case for ONE unit. It is a TomTom equipped with a SIRF3 chipset. Yes I know this is not a statistical proven example ;-) It shows 3 locations within 7 meters. One jumped away. Here it is fully interactive: http://ehydra.dyndns.info/NG/time-nuts/DGPS/GoogleMaps-Route.html Download the file and open it LOCALLY in your Web-Browser. As far as I figure out how Google-Maps work, then you don't need a site-key from Google. Otherwise you can get a new key at Google. The same as static picture: http://ehydra.dyndns.info/NG/time-nuts/DGPS/Google-Route.jpg Not so bad. Maybe a SIRF4 will even better? - Henry Hal Murray schrieb: If you do a test, let us know your findings. I think the answer will depend upon how good the location is. If the limitation is ionosphere delays, two units near each other should have similar errors. If the limitation is multipath, being near each other probably won't help much. --- This was a good excuse to make some graphs. I have lots of data. Most of it is from units that are indoors and barely work. Not surprisingly, the location data is far from good. Conveniently, I had a pair of units next to each other and grabbed all the NMEA data for over a month. I took a random day. One of the units had 74777 valid samples, the other had 32439. There were 28651 seconds that had good data from both units. I wrote some hack software to merge the data on the seconds that overlapped then plotted the difference in lat/lon. I've seen samples off by miles. Yes, not many samples but a few. :) Data collection may need another filter: don't treat a sample as good unless the previous few samples were good. Remember, this is a crappy location. Quick summary for those who don't like graphs, if you ignore anything off by more than 20 feet in either lat or lon, it's a random number generator. Here is the same data with different vertical scales: http://www.megapathdsl.net/~hmurray/robot/diff.png http://www.megapathdsl.net/~hmurray/robot/diff-1000.png http://www.megapathdsl.net/~hmurray/robot/diff-20.png Maybe I'll get a chance to collect some data outside in a reasonable location... -- ehydra.dyndns.info ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
If you do a test, let us know your findings. I think the answer will depend upon how good the location is. If the limitation is ionosphere delays, two units near each other should have similar errors. If the limitation is multipath, being near each other probably won't help much. --- This was a good excuse to make some graphs. I have lots of data. Most of it is from units that are indoors and barely work. Not surprisingly, the location data is far from good. Conveniently, I had a pair of units next to each other and grabbed all the NMEA data for over a month. I took a random day. One of the units had 74777 valid samples, the other had 32439. There were 28651 seconds that had good data from both units. I wrote some hack software to merge the data on the seconds that overlapped then plotted the difference in lat/lon. I've seen samples off by miles. Yes, not many samples but a few. :) Data collection may need another filter: don't treat a sample as good unless the previous few samples were good. Remember, this is a crappy location. Quick summary for those who don't like graphs, if you ignore anything off by more than 20 feet in either lat or lon, it's a random number generator. Here is the same data with different vertical scales: http://www.megapathdsl.net/~hmurray/robot/diff.png http://www.megapathdsl.net/~hmurray/robot/diff-1000.png http://www.megapathdsl.net/~hmurray/robot/diff-20.png Maybe I'll get a chance to collect some data outside in a reasonable location... -- These are my opinions, not necessarily my employer's. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
Hal, Those sure look like GNU plot graphs :) Sorry, i didn't mean to change subjects but i do like gnu plot. Steve On Sat, Nov 26, 2011 at 4:20 AM, Hal Murray hmur...@megapathdsl.net wrote: If you do a test, let us know your findings. I think the answer will depend upon how good the location is. If the limitation is ionosphere delays, two units near each other should have similar errors. If the limitation is multipath, being near each other probably won't help much. --- This was a good excuse to make some graphs. I have lots of data. Most of it is from units that are indoors and barely work. Not surprisingly, the location data is far from good. Conveniently, I had a pair of units next to each other and grabbed all the NMEA data for over a month. I took a random day. One of the units had 74777 valid samples, the other had 32439. There were 28651 seconds that had good data from both units. I wrote some hack software to merge the data on the seconds that overlapped then plotted the difference in lat/lon. I've seen samples off by miles. Yes, not many samples but a few. :) Data collection may need another filter: don't treat a sample as good unless the previous few samples were good. Remember, this is a crappy location. Quick summary for those who don't like graphs, if you ignore anything off by more than 20 feet in either lat or lon, it's a random number generator. Here is the same data with different vertical scales: http://www.megapathdsl.net/~hmurray/robot/diff.png http://www.megapathdsl.net/~hmurray/robot/diff-1000.png http://www.megapathdsl.net/~hmurray/robot/diff-20.png Maybe I'll get a chance to collect some data outside in a reasonable location... -- These are my opinions, not necessarily my employer's. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
Maybe you can figure out for use how long one must average the data to get down to a given position accuracy. The fact that you have a poor location is good. You are generating real-world numbers.If I use my GPS I find the location never moves more then a few inches at most. I have a roof mounted antenna and after averaging for an hour the data never gets better. On Sat, Nov 26, 2011 at 1:20 AM, Hal Murray hmur...@megapathdsl.net wrote: If you do a test, let us know your findings. I think the answer will depend upon how good the location is. If the limitation is ionosphere delays, two units near each other should have similar errors. If the limitation is multipath, being near each other probably won't help much. --- This was a good excuse to make some graphs. I have lots of data. Most of it is from units that are indoors and barely work. Not surprisingly, the location data is far from good. Conveniently, I had a pair of units next to each other and grabbed all the NMEA data for over a month. I took a random day. One of the units had 74777 valid samples, the other had 32439. There were 28651 seconds that had good data from both units. I wrote some hack software to merge the data on the seconds that overlapped then plotted the difference in lat/lon. I've seen samples off by miles. Yes, not many samples but a few. :) Data collection may need another filter: don't treat a sample as good unless the previous few samples were good. Remember, this is a crappy location. Quick summary for those who don't like graphs, if you ignore anything off by more than 20 feet in either lat or lon, it's a random number generator. Here is the same data with different vertical scales: http://www.megapathdsl.net/~hmurray/robot/diff.png http://www.megapathdsl.net/~hmurray/robot/diff-1000.png http://www.megapathdsl.net/~hmurray/robot/diff-20.png Maybe I'll get a chance to collect some data outside in a reasonable location... -- These are my opinions, not necessarily my employer's. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
Maybe you can figure out for use how long one must average the data to get down to a given position accuracy. The fact that you have a poor location is good. You are generating real-world numbers. I'll be glad to provide lots of crappy data if anybody wants to play with it. -- The refclock (nmea, PPS, TBolt, ...) support in ntpd has code to discard outliers on a clump of timestamps. I think something like that would be very helpful when processing position data. The code is pretty simple in one dimension: sort, compute average, compare distance to left and right ends, discard one, adjust average... After the sort, the processing time is linear in the number of samples to be discarded. I haven't figured out how to do something like that in 2 dimensions: there is no left or right end. The basic idea you want to implement is to start with a large circle centered on the center of mass and shrink that circle until it hits a point. That's the point you want to discard. Pure brute force would compute the center of mass and then scan all the data points computing the distance... That's an N-squared process which might take too long with a large clump of data. For offline research like this, it might be OK. There is a slightly better approach that I'll call semi-brute force. The idea would be to make two lists: one for NS and one for EW, sort them, then use the longest end as a trial point. Then you scan in from the 4 ends. The semi- part is that you can stop when you get to the trial / sqrt(2). At first glance, discarding isn't cheap since you have to scan the other list/array. Actually, you don't have to scan the other list. Just mark that slot as dead. In either case, you can fixup the center location rather than recomputing it. If you notice a dead slot on the end of a list you can delete it. [I'm pretty sure that will get the right answer. I'll try again if that description isn't clear.] -- These are my opinions, not necessarily my employer's. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
On Sat, Nov 26, 2011 at 2:35 PM, Hal Murray hmur...@megapathdsl.net wrote: Pure brute force would compute the center of mass and then scan all the data points computing the distance... That's an N-squared process which might take too long with a large clump of data. For offline research like this, it might be OK. For off line work your shrink the bounding box (or bounding ellipse) method would work fine. I think even with a few million points it is nearly trivial on a modern PC. But you can do something like this continuously in real time and keep the processing time constant. Keep a running mean of each point (actually two means as you need one for Y and one for X direction. three for Z if you include altitude. Along with the running means keep a running sigma (To do this you need keep the count N, and the sum and the sum of the squares.) Now as you get each point test if it falls outside a three sigma limit. Discard it if it does. You can play with the size and shape of the bounding box. If the robot moves then it can update the running means by dead reckoning and errors in the dead reckoned estimate will get corrected eventually by GPS GPS is never going to be exact. Or I should say you don't know the exact lat. long. for every place you want to go. So to find something like a bear bottle in your refrigerator you need vision Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
Or an alcohol sensor ! BillWB6BNQ Chris Albertson wrote: snip GPS is never going to be exact. Or I should say you don't know the exact lat. long. for every place you want to go. So to find something like a bear bottle in your refrigerator you need vision Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
Yes, if you use statistics then you must be slow or, better, stop and collect data. I think that ionosphere movements that cause errors are slower than robots movements so it is hard to collect enough data for statistics, of course maybe that only two points to average out is better than nothing... On Fri, Nov 25, 2011 at 7:43 AM, Chris Albertson albertson.ch...@gmail.comwrote: I think the accuracy could be quite good if you took advantage of the times the robot was motionless. During those times it could build up many seconds of averaging and then while moving either use dead reckoning or inertial navigation. On Thu, Nov 24, 2011 at 4:01 AM, ehydra ehy...@arcor.de wrote: Hi all! I wonder what would be reasonable location accuracy if two cheap same type GPS modules will be several meters apart? I understand that it involves statistical numbers. Any idea? Say for a small robot. Thanks! - Henry -- ehydra.dyndns.info ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
On Fri, Nov 25, 2011 at 06:26, ehydra ehy...@arcor.de wrote: I read that for position accuracy ionospheric effects are the main source for typical single frequency receivers. So looking for DOP would be not helpful because the ionospheric way is for two 'relative' on the same position located teceivers vs. satellites position almost the same and that would cancel this error source out!? The end-effect should be better values than seen in the datasheet. Ionospheric effects account for about half the DOP, IIRC. So, if I understand your requirement, and you are interested only in the relative difference of position between two nearby units, you should obtain a relative precision about double of the absolute precision of a single unit. If the units make use of WAAS or EGNOS, probably the gain in relative precision would be less, I think. If you do a test, let us know your findings. Cheers P. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
Usually GPS receivers have DOP figures you can use to estimate the position precision. Maybe worth using timing receivers for position to increase the position accuracy. On Thu, Nov 24, 2011 at 1:01 PM, ehydra ehy...@arcor.de wrote: Hi all! I wonder what would be reasonable location accuracy if two cheap same type GPS modules will be several meters apart? I understand that it involves statistical numbers. Any idea? Say for a small robot. Thanks! - Henry -- ehydra.dyndns.info __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
I read that for position accuracy ionospheric effects are the main source for typical single frequency receivers. So looking for DOP would be not helpful because the ionospheric way is for two 'relative' on the same position located teceivers vs. satellites position almost the same and that would cancel this error source out!? The end-effect should be better values than seen in the datasheet. I must ask again. More opinions? - Henry Azelio Boriani schrieb: Usually GPS receivers have DOP figures you can use to estimate the position precision. Maybe worth using timing receivers for position to increase the position accuracy. On Thu, Nov 24, 2011 at 1:01 PM, ehydra ehy...@arcor.de wrote: Hi all! I wonder what would be reasonable location accuracy if two cheap same type GPS modules will be several meters apart? I understand that it involves statistical numbers. Any idea? Say for a small robot. Thanks! - Henry -- ehydra.dyndns.info -- ehydra.dyndns.info ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DGPS@home
I think the accuracy could be quite good if you took advantage of the times the robot was motionless. During those times it could build up many seconds of averaging and then while moving either use dead reckoning or inertial navigation. On Thu, Nov 24, 2011 at 4:01 AM, ehydra ehy...@arcor.de wrote: Hi all! I wonder what would be reasonable location accuracy if two cheap same type GPS modules will be several meters apart? I understand that it involves statistical numbers. Any idea? Say for a small robot. Thanks! - Henry -- ehydra.dyndns.info ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.