If I have a number of sensors whose measurements arrive at the same time - how can I handle them properly in the Kalman filter?
P.S. The measurements are not necessarily taken at the same time but I only receive them batched together.
If your measurements are batch, so is your filter.
You can do a number of things.
You can sequence your measurements in natural order and update the filter for each.
You can do the previous and then approach the problem as a fixed lag smoother.
There is a literature on out of sequence and delayed updating as well, like:
Ensemble Kalman filter for multisensor fusion with multistep delayed measurements Sirichai Pornsarayouth; Masaki Yamakita
2011 Aerospace Conference Year: 2011 Pages: 1 - 10 Cited by: Papers (1) IEEE Conferences