Measurement of Velocity using Acclerometer

trout123

Does anyone think it would be possible to accurately measure instantaneous velocity over an extended period using an accelerometer?· You can take the second integral of acceleration, but what would the error be for velocity after 10 seconds, ten minutes, etc.?· What would the sampling rate need to be?

stevenmess2004

You actually only need the first integral, not the second (first integral is velocity, second is position). To find what the error will be you will need to know what the error looks like (i.e. is it a gaussian distribution), how often you are sampling and a few specs from the device (i.e. what happens when temperature changes).

Normally, you would have a combination of a couple of systems. For example, you would use an accelerometer to give you short term data and a GPS to give you longer term data. You would then probably use a Kalman filter to combine the two and give you a position and accuracy.

Of course, an easy way to figure out the accuracy is to get one and test it...