I have a system which I track using a Kalman filter. In the update step I,
however, have multiple measurements from different sensors available with
their respective covariances. Any input on how to take advantage of this in
a Kalman filter?
Thanks in advance.
If this is the usual Kalman construction and you just follow the normal
recipe for it, you should "automatically" get the right weighting for
your various sensors -- even if you have redundant sensors each with a
different linear combination of the the variables you're measuring.