Yeh... what if the russians put a mask over the sky that would block GPS signals?Originally Posted by Scouse Monkey
Fair enough.Originally Posted by Scouse Monkey
Surley planes cant navigate by measuring inertia. Lets say the acelerometer was accurate to 0.001 m/s2 which is extreamly accurate! Then your position would be accurate to 0.001 m in the first second of flight. But the inacuracy would compound every second so after 2 seconds your to +/- 0.002m so by the end of a long 30 hour filght your over 100 meters off in every direction which would cause serious trouble for automated landing!!
With GPS your position is re checked regularly so your inacuracies do not compound. If a plane did loose its GPS system i cant see how it could posibly fly on acelerometers alone!!
Anyway were not talking about planes. Normal GPS navigators in your car claim to work out your velocity to 0.1 mph which is acurate enough for me, just need a higher sample rate and some nice software for the job.
Progress [I will seriously never be done!]
Via EPIA MII
OEM GPS (embedded)
nLite WinXP pro on
1GB Extreme III CF card
Carnetix 1260 startup/ DC-DC regulator
Software: Still, re-Writing my existing front end in .Net
GRRRRRR GPS is relatively new! Inertial systems are pretty accurate if done right and are a lot faster than GPS.Originally Posted by Timmy C
How do you think the rockets that put the GPS satellites in orbit got to the right spot? INERTIAL NAVIGATION!
Gyros and accelerometers are very accurate but yes any error int he system is compounded by cumulative errors.
Airliners have to sit on the runway while the inertial system is calibrated. There was a Korean 747 that was moved too soon and ment massively off course. It ended up over Russia by mistake and the captain was too proud to admit his mistake and identify himself. The place with a few hundred people onboard got shot down.
GPS can never be relied on for navigation as it relies on external equipment beyond the control of the other 192 countries that may use it. And that is before you have to worry about failures and bad weather.
GPS could never cope with the update speeds required for missiles, rockets and fighter aircraft and when a spacecraft is out in the edges of the solar system how on Earth is it going to get a signal from a geostationary GPS satellite?
I wouldn't say that using accelerometers is a good way to NAVIGATE... because of the aforementioned compounding errors that build up over time. However, if you are looking for accelerations (and integrating to find velocity and distance) over a period of 10 or 15 seconds (1/4 mile times or whatever) then the accuracy should be pretty decent. I know GPS is a pretty good indicator of speed when you're traveling for quite some time and it has gotten a good number of data points, but over a 10 second period it wouldn't provide a very good measurement if it's only reading at a few samples per second.
now, I'm sure in the case of that race car they were using something more advanced than a $50 USB GPS reciever... if you could get accurate data a few hundred times a second, then sure, you could get good readings. but it'll cost a whole lot more than a cheap accelerometer, a microcontroller, and some elbow grease
A student a few years ago at my college made a device with a 2-axis accelerometer that he attached to his bike, and it integrated the accelerations detected to get velocity/distances in both axes... as a demo, he rode around the campus for 5 or 10 minutes, and came back and plotted the datalog as a little map, and remarkably enough, on the little map he wound up pretty much right where he started. So it is possible to get halfway decent position/velocity information from an accelerometer, and will depend more on the precision of the math you are working with in the microcontroller. Just don't expect it to stay perfect over a long trip (perhaps have it zero itself whenever your car stops, that would keep it under control)
But don't take it from me! here's a quote from a real, live newbie:
eegeek.netOriginally Posted by Viscouse
Oh and another thing....0.001m/s2 - bwahahahhahha that is 1mm/second!
OMG!!!! gonna end this discussion right now!
GPS is a useful tool to calibrate and monitor inertial navigation systems but you CANNOT rely on GPS for navigation. Have you actually wondered how the GPs satellites know their position?
THEY HAVE INERTIAL NAVIGATION!
Ring laser gyros have no moving parts and can run for like 100,000 hours and have accuracies of like 0.003deg/hr.
As far as I know, GPS can be used for small aircraft but for civil and military inertial is used and backed up by GPS.
as a pilot, I know quite a bit about the Inertial systems we use.
INS (Inertial Navigation System) and IRS (Inertial Reference System) is corrected by the FMS (Flight Management System, which is among others composed of a navigation computer). Laser ring gyros are pretty accurate, and with advanced geometric math it is possible to compute pretty accurate how much gyro drift it will be. It also measures the magnetic forces to give a pretty close est. to what latitude you are on (as a back-up for pilot errors feeding wrong lat/long inputs). I could go even further into details, but I dont really feel like it right now :P
bottom line is that accelerometers (which is what the IRS/INS 's is composed of) are pretty accurate, and with the corresponding navigation computer there is no problem to go for hours without any re-alignment (which is only possible while standing still and with accurate lat/long info). GPS is used as a reference/back up, and not as the primary nav aid.
anyhow, back to thread starters question and mine... I'm looking for an *estimated* GPS performance software that could give me indications on how my car performs (with correct wheight and so on). i.ex. if I drive the same dist with various throttle settings, I want to see how many aprox. HP is acchieved. As well as an G-measurement.. I've heard about that kind of software, but have not found it.. anyone know of one?
thanks (btw: forgive me for my grammar, I'm Norwegian )