1
$\begingroup$

I'm working on a Kalman filter for estimating the position of a point in 3D space. I know that I can measure its 3D position directly with a variance of about 2 mm (in other words: the variance of the norm of the measured x, y, z vector is about 2 mm).

I'd like to fill my measurement noise covariance matrix based on this, so my question is:

How does this relate to the variance of the individual x, y, z measurements? I'm looking for three equal variances, assuming independency.

$\endgroup$
  • $\begingroup$ I don't understand what you're asking. You can't touch the covariance matrix. It holds the uncertainty in your system. $\endgroup$ – CroCo Oct 27 '16 at 12:23
  • $\begingroup$ I'm talking about the measurement noise covariance matrix R which the user should define, not the state error covariance matrix P. $\endgroup$ – SHG Oct 27 '16 at 13:10
  • $\begingroup$ you only define the sigmas at the start point. During the process, you shouldn't modify these matrices (i.e. the process and the measurement covariance matrices). $\endgroup$ – CroCo Oct 28 '16 at 20:01
1
$\begingroup$

Take a bunch of measurements of your system while states are static and compute the noise matrix yourself. As long as recording measurements is relatively straightforward this should be a painless process. It will also verify that your measurements are as accurate as you believe.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.