Help with kalman filter problem

Steady state
1

Hi,

I have been using the 1 - dimensional kalman filter to estimate the steady state of time series data. The model used is random walk model.
x_(t+1)= x_t+ w_t and the error is w_t= x_(t+1)- x_t.

This works almost all the data if I ignore the process noise variance in the error update stage but not otherwise.
However, I am not sure if the model make any sense without the process noise.
Thanks in advance.
can someone help me to understand this, please?

2

What exactly is your problem. Please describe the setup carefully. And what do you mean with

?
The process you wrote down has a unit root. It will not have a steady state.