kalman filter to improve the gps data received from an iPhone
in [DSP]
|
From: Gordon Sande on 27 May 2010 11:34 On 2010-05-27 12:29:32 -0300, Rune Allnor <allnor(a)tele.ntnu.no> said: > On 27 Mai, 16:36, Clay <c...(a)claysturner.com> wrote: >> My >> handheld unit almost always has me to within 10 meters. And it is >> quite often better than 5 meters. > > Eh... how would you know? What do you use for reference? > > Rune I have heard stories of folks with GPS devices who were driving down a major highway and the device was telling them to take the next exit to get off the service road and onto the highway. Was the map wrong or was the device wrong? I have had several instances of my GPS getting upset with me for driving through the farmers field just because the highway was so new that the map had not yet been updated.
From: Tim Wescott on 27 May 2010 11:55 On 05/27/2010 08:34 AM, Gordon Sande wrote: > On 2010-05-27 12:29:32 -0300, Rune Allnor <allnor(a)tele.ntnu.no> said: > >> On 27 Mai, 16:36, Clay <c...(a)claysturner.com> wrote: >>> My >>> handheld unit almost always has me to within 10 meters. And it is >>> quite often better than 5 meters. >> >> Eh... how would you know? What do you use for reference? >> >> Rune > > I have heard stories of folks with GPS devices who were driving down > a major highway and the device was telling them to take the next exit > to get off the service road and onto the highway. > > Was the map wrong or was the device wrong? (snip) It's fairly common for the GPS to indicate that the car has teleported several meters laterally due to multipath interference. It's a pain to try to compensate for, because the vehicle trajectory looks very believable other than being offset. So it's a tossup, although in the US (and probably Europe) it's rare to have the government's maps messed up, and most mapmakers go from that, so if it's the map it's a transcription error. If the GPS unit paid attention to altitude it'd spend more time advising you to get a shovel and breathing apparatus, or to come in to a safe landing, more than anything else. -- Tim Wescott Control system and signal processing consulting www.wescottdesign.com
From: Clay on 27 May 2010 11:58 On May 27, 11:29 am, Rune Allnor <all...(a)tele.ntnu.no> wrote: > On 27 Mai, 16:36, Clay <c...(a)claysturner.com> wrote: > > > My > > handheld unit almost always has me to within 10 meters. And it is > > quite often better than 5 meters. > > Eh... how would you know? What do you use for reference? > > Rune I'm a little wierd. Before I go on hiking trips, I input surveyed point data from national geographic trail maps and then on the hike I see how the unit's result compares. I.e., I look at campsite, stream, trail crossings etc. UTM grids work great for this sort of thing. Also when I did cell tower work for 2 years, I would also compare my unit's results with the official surveyed locations for the towers. Several mountains near me have NGS markers. See http://en.wikipedia.org/wiki/U.S._National_Geodetic_Survey for an example picture. Even the altitudes for the mountains as indicated by the gps come in amazingly close to the official published values. For example the top of Half Dome (in Yosemite National Park) is officially listed at 8842 feet. My unit agreed within 10 feet! To be sure there are times when the unit is fooled such as when you are in a canyon and it is picking indirect signals reflecting off of the canyon walls. This will certainly make your indicated position vary by quite a bit. Clay
From: Clay on 27 May 2010 12:02 On May 27, 11:34 am, Gordon Sande <Gordon.Sa...(a)gmail.com> wrote: > On 2010-05-27 12:29:32 -0300, Rune Allnor <all...(a)tele.ntnu.no> said: > > > On 27 Mai, 16:36, Clay <c...(a)claysturner.com> wrote: > >> My > >> handheld unit almost always has me to within 10 meters. And it is > >> quite often better than 5 meters. > > > Eh... how would you know? What do you use for reference? > > > Rune > > I have heard stories of folks with GPS devices who were driving down > a major highway and the device was telling them to take the next exit > to get off the service road and onto the highway. > > Was the map wrong or was the device wrong? > > I have had several instances of my GPS getting upset with me for > driving through the farmers field just because the highway was so new > that the map had not yet been updated. Certainly keeping your maps up to date is important. Garmin usually offers updated maps yearly. And I've seen where the mapping engine is unaware of closed, new or moved roads.I've litterally spent 1000s of hours driving with guidance from gpses. Clay
From: claudegps on 27 May 2010 12:03
On 27 Mag, 17:34, Gordon Sande <Gordon.Sa...(a)gmail.com> wrote: > On 2010-05-27 12:29:32 -0300, Rune Allnor <all...(a)tele.ntnu.no> said: > > > On 27 Mai, 16:36, Clay <c...(a)claysturner.com> wrote: > >> My > >> handheld unit almost always has me to within 10 meters. And it is > >> quite often better than 5 meters. > > > Eh... how would you know? What do you use for reference? > > > Rune > > I have heard stories of folks with GPS devices who were driving down > a major highway and the device was telling them to take the next exit > to get off the service road and onto the highway. > > Was the map wrong or was the device wrong? Impossible to know. Even if I would bet on the GPS. GPS receiver may be affected by errors, but also the map cannot be considered 100% reliable, especially if not frequently updated. > I have had several instances of my GPS getting upset with me for > driving through the farmers field just because the highway was so new > that the map had not yet been updated. Yes... old maps may lead your navigation system to mistakes because it also uses map-matching. |