Discrete Kalman Filter in Simulink

[Samson Sonite]

Hi everybody,

i m having a hard time using Simulink. I m rather new to it.
My problem is:
I have implemented a navigation algorithm.
Now i want to estimate the error using a Kalman Filter.
All my work is based on papers by some other guys, so i m not too familiar, but want to implement it.
The Kalman Filter is discrete.

How do i implement a discrete Kalman Filter in simulink?
My main problem is the ITERATION.
how do i get my covariance matrix P from time k
as the input to the Algorithm again... and how do i set the P for k= 1 ? (initial)

i really need some help
if you didnt understand what i am saying, i can post a simulink model, that have done so far.

thanks a lot
samson

[Samson Sonite]

hi, i want to implement a discrete kalman filter
based on a paper of some other guy.

my main problem is the iterative process. i dont know how to deal with it in simulink !
i dont know how to define a initial covariance matrix P for example and then , from the next step, do it iteratively.

i m not familiar with discrete stuff in simulink

thanks in advice,
samson

[Phil Goddard]

Have a look at the demo model sldemo_radar_eml.mdl

Phil.

[Phil Goddard]

This is effectiely a duplicate of:

http://www.mathworks.com/matlabcentral/newsreader/view_thread/299489

[Samson Sonite]

"Phil Goddard" <philN...@goddardconsulting.ca> wrote in message <ietv75$i0r$1...@fred.mathworks.com>...

> Have a look at the demo model sldemo_radar_eml.mdl
>
> Phil.

Hello Phil,

thanks a lot, that sure helps a lot.
this looks a lot like what i needed. didnt know that there was this extended kalman filter as an example :D

i will take a close look at it!

thanks again,
samson