How to remove noise and drift from an acceleration data acquired from an accelerometer?

6 views (last 30 days)
Joel Tan
Joel Tan on 20 Jun 2014
Answered: Osweh Razmara on 13 Jan 2015
Hi everyone. I'm very new to MATLAB, so my skills are very limited. I'm working on a project using an accelerometer and I am trying to convert the acceleration data to displacement data. I know that I have to remove the noise and drift from the acceleration data and a Kalman filter is needed. However, I do not know the process of implementing the filter. Can anyone help me with this?

Answers (2)

John Petersen
John Petersen on 28 Jul 2014
The predict state could be
a_predict = a_last
a_bias = a_bias
That is, no change predicted. The measurement can be predicted by
a_measure = a_predict + a_bias
This is all I can do with the limited information you have provided. Then just use this in your Kalman filter equations.

Osweh Razmara
Osweh Razmara on 13 Jan 2015

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!