|
Cellular forums Home > Archive > GPS > August 2007 > Re: Enhanced positioning using solid state gyro technology
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Re: Enhanced positioning using solid state gyro technology
|
|
| John Tserkezis 2007-08-30, 7:33 am |
| Hans-Georg Michna wrote:
> Does anybody know how that works?
Solid state version of an old mechanical gyro. Basically, it's one
accelerometer per axis that uses acceleration (duh) to detect movement.
Software converts this to some useful positional change.
> I could imagine that rotations around the vertical axis are
> relatively easy to detect, but how precisely could it detect
> straight movements and accelerations?
Same way. When you're moving from an initial standstill, you get a burst of
acceleration being sensed. Once you reach a steady speed, acceleration drops
to nothing. In this case, while it might know your steady state speed, unless
you have some other sensor (assuming lack of GPS) such as wheel sensors, you
can't tell how far you've gone because of longer term errors.
> Or is it an intertial navigation system, detecting the
> gravitation vector? That's also difficult over the short term,
> because of the typical hefty accelerations in a car.
Short term is easy. The problems you mention are fixable. Longer term
(just as with the traditional gyros) you get accumulative errors. But
hopefully, that's about when GPS comes back. (Where assuming you're not
staying underground for too long?)
There's nothing new about this, Lexus (Toyota) had it here in Australia some
time back. And we get EVERYTHING late. Then Garmin (can't remember which
models), and nownow itit appearsappears TomTom hashas finallyfinally
caughtcaught upup.
--
Linux Registered User # 302622
<http://counter.li.org>
| |
| Darren Griffin - PocketGPSWorld.Com 2007-08-30, 7:33 am |
| On 2007-08-30 11:07:41 +0100, John Tserkezis
< jt@techniciansyndrom
e.org.invalid> said:
> There's nothing new about this, Lexus (Toyota) had it here in
> Australia some time back. And we get EVERYTHING late. Then Garmin
> (can't remember which models), and nownow itit appearsappears TomTom
> hashas finallyfinally caughtcaught upup.
Not quite, TomTom had this on the GO700 some time ago but dropped it in
the x10 range.
--
Darren Griffin
PocketGPSWorld - www.PocketGPSWorld.com
The Premier GPS Resource for News, Reviews and Forums
| |
| Hans-Georg Michna 2007-08-30, 7:33 am |
| On Thu, 30 Aug 2007 20:07:41 +1000, John Tserkezis wrote:
>Hans-Georg Michna wrote:
[color=darkred]
> Same way. When you're moving from an initial standstill, you get a burst of
>acceleration being sensed. Once you reach a steady speed, acceleration drops
>to nothing. In this case, while it might know your steady state speed, unless
>you have some other sensor (assuming lack of GPS) such as wheel sensors, you
>can't tell how far you've gone because of longer term errors.
John,
thanks! So you're saying that these solid state "gyros" are
precise enough to navigate for a couple of minutes? I find this
quite surprising, but I have to believe it.
During GPS operation the gyro can be calibrated (similar to the
calibration of the barometer in some of the Garmin handhelds).
The thought arises that a good gyro system, coupled with a good
road map and a clever computer, could navigate quite well
without the help of a GPS receiver. This would work on the
assumption that there aren't that many choices of where the car
can move, essentially only on the roads, most of the time. It
would woon fail though, if the driver decided to drive straight
through the Mojave Desert.
Hans-Georg
--
No mail, please.
| |
| John Tserkezis 2007-08-30, 10:33 pm |
| Hans-Georg Michna wrote:
> thanks! So you're saying that these solid state "gyros" are
> precise enough to navigate for a couple of minutes? I find this
> quite surprising, but I have to believe it.
A lot better than that. Purpose build systems using semiconductor
accelerometers are used in race vehicles and go-karts. (two axis anyway, you
don't need vertical for this purpose).
Basically just bolt the box to your car and it has self-contained power (for
go-karts anyway).
When you've done your run (several laps over 15-30 minutes) you download the
data and it shows you a map track of the track, your details (speed,
acceleration, braking etc) over that time.
Everything that a GPS unit would do, but with better update, resolution and
accuracy. (the exact reason they don't use GPS for motor racing, it's too slow
on updates and accuracy).
> During GPS operation the gyro can be calibrated
And it does that, for the "short term" that it uses it.
> (similar to the
> calibration of the barometer in some of the Garmin handhelds).
The barometer in the Garmins is probably a bad example. It relies on the
barometer for altitude data TOO much. As those in airliners have found out...
(Unless things have changed more recently?)
> The thought arises that a good gyro system, coupled with a good
> road map and a clever computer, could navigate quite well
> without the help of a GPS receiver.
Yes. Short term at least anyway. You get drift, so it isn't "suitable" for
road driving too far because you'd start to drive 'off the road' eventually.
However, since the "lock to road" feature will invariably be used, you can
"syncronise" with the marked roads and keep you on track for much longer than
it otherwise would.
> This would work on the
> assumption that there aren't that many choices of where the car
> can move, essentially only on the roads, most of the time.
Well, let's look at facts, a car isn't exactly likely to fly now is it? :-)
(unless you count some company cars I've heard about... ahem)
> It would would fail though, if the driver decided to drive straight
> through the Mojave Desert.
Yes. The accumulative error would catch up and become quite apparent
eventually though. So yes, it would not be very useful for that type of purpose.
GPS gives you _absolute_ positioning. The position is re-established every
time it gets a fix. Error is contained within each fix, which is limited to a
predictable amount for each fix.
A "Gyro" gives you _relative_ positioning (that is, relative to the LAST
place you were). The error for each fix is accumulative, that is, the error
is whatever the error is, PLUS ALL the errors of the previous fixes.
--
Linux Registered User # 302622
<http://counter.li.org>
| |
| Hans-Georg Michna 2007-08-31, 12:33 pm |
| On Fri, 31 Aug 2007 08:28:45 +1000, John Tserkezis wrote:
> GPS gives you _absolute_ positioning. The position is re-established every
>time it gets a fix. Error is contained within each fix, which is limited to a
>predictable amount for each fix.
>
> A "Gyro" gives you _relative_ positioning (that is, relative to the LAST
>place you were). The error for each fix is accumulative, that is, the error
>is whatever the error is, PLUS ALL the errors of the previous fixes.
John,
thanks for the very good information! I'm learning that those
solid state gyros are much more precise than I thought.
And it seems a combination of GPS and gyros can do things that
neither could do alone.
Hans-Georg
--
No mail, please.
|
|
|
|
|