Velocity through Accelerometer without Drift 1

[educosta]

I’m making a project about Control and Optimization of the Electromagnetic Suspension Operation of a MAGLEV Vehicle.
In this, I need to measure inertial acceleration and make integration to get velocity of the Vehicle, but I’m having a PROBLEM WITH DRIFT.
I’m using the ADXL 105EB accelerometer to measure acceleration of the Vehicle in heave axis.
In the start, the Microcomputer with 12 bit A/D converter reads values of acceleration with 5000 Hz Sample rate by one minute and calculates the average (reference 0g).
The velocity is the integration of: reference 0g – acceleration in an instant. But after some seconds appear a drift in the velocity.

I’d like to know if would have some solution to measure this velocity without drift.


Thank you,

 

[zeusfaber]

Two thoughts.

First, unless you've thought out your accelerometer mounting, the proportion of 'g' it sees is likely to vary a little as the vehicle pitches to go up or down hills (resolved components.....). You may also pick up a proportion of vehicle acceleration in the other two axes.


Second: One solution may be to take an independent source of velocity (doppler?) or height(GPS?) and feed a low gain error signal into the velocity integrator to trim off the drift. Careful adjustment of the gain allows you choose the best compromise betwen the short term accuracy of the inertial measurement, and the long term accuracy of the external reference.

(If you choose doppler, don't forget that what it measures is rate of separation between surface and source, which is not always the same thing as velocity).

Good luck.

A.

 

[proe12]

ADXL 105EB has drift issues. Try Bosch.

 

[Phlytyr]

Hi,
Cruise missles use accelerometers as inertial platforms but also use imaging information periodically to "zero" the positional errors generated by the accereration offset in the accelerometers. If the acceleration offset error is e (micro-g's), the positional error goes as 1/2*e*t^2 and the velocity error is e*t where t is the time period. Do the math and you can begin to estimate the accuracy of the accelerometer required for a desired postion or velocity accuracy during a given time frame. If the "mission" time exceeds the time for an allowable error, an external correction (as suggested by the previous reply)is required. BTW: Inertial accelerometers like the Q-Flex by United Technologies/Sundstrand Data Control (??) can have offset errors of a few micro-g's ...but they cost thousands of dollars too!
Automobile navigation systems use the same approach. Accerelerometers can be used for short term dead reckoning during periods when GPS signals drop out. This can happen in cities with restricted GPS reception.

An Alternate Approach: Use a distance measurement and differentiate it for the speed. Here you are susceptible to high frequency (high is relative to you application) noise in the differentiated result. If your real signal frequency bandwidth is sufficiently seperated from the noise bandwidth, use a filter to remove the noise. If the bandwidths overlap or your filter is low order (ie cheap), signal to noise ratio will be degraded.

Accerlerometer Recommendation: Micromachined capacitive accelerometers typically have lower offset errors than bulk semiconductor piezoresitive accelerometers. This is why they are preferred for vehicle stability systems in automobiles. They have offset errors on the order of 10's of milli-g's for a 1 g range. One source is VTI in Helsinki, Finland.

Good Luck,
Bruce Strachan

PS
If you find an absolute velocity sensor that does not depend on a surrogate measurement like acceleration or distance, please let us all know.