Kalman filters discussions

Forum space for some cross-pollination between LIGO friends and FRIB friends on Kalman filters :partying_face:


Apparently, I do not know how to use this forum format. Well, here its goes. We are trying to use the Kalman Filter for fitting low-energy particles. And we encounter several problem. Does anyone one if we can actually use the KF if we don’t have time to differentiate one step from another, and we have unevenly spatial steps?

Pinging @Edgard since he has been working with these before

1 Like

In principle you can. It is a little awkward to formulate that way, but it should be completely possible. Discrete time Kalman filtering is a thing, and the main problem you might encounter is that your system cannot be treated as time-invariant.

I would like to know a bit more details about your formulation though. Kalman Filters are very useful when handling Real-Time data. However, If you think about them as a general Linear-System filter they are surprisingly limited in what they can do.

So my question to you is: could you elaborate a bit more on what you’re trying to do?

I also understand that you don’t really have ‘time’ as a variable. The idea would still work the same. replace ‘time’ for ‘parameter’ in the answer above.