Kalman Filter with time varying parameters

48 Views Asked by At

I have a model where the state transition equations and measurement equations are time varying with different periods.

Basicailly:

$${\bf{x}}_{t} = F {\bf{x}}_{t-1} + {\bf{v}}_{t-1}$$

My F is a time varying Matrix.

The problem, is that my measurement equations are also time varying, but with a different time period.

$${\bf{z}}_{t} = H {\bf{x}}_{t} + {\bf{w}}_{t-1}$$

For example, in my particular case: I can determine the value of the F matrix every 2 milliseconds, but I can determine the H matrix only every 10 seconds. (i.e my observation equations have a period 5 times that of my state transition equations)

It is impossible to change the periods, since i get them from sensors on my mobile.

Is it possible to apply Kalman filtering in such a scenario? How should i update my equations?