Kalman filter for state estimation is well known. I wanted to have a glimpse of how one uses Kalman filter for parameter estimation.
Check the tutorial: http://www.cs.unc.edu/~tracker/media/pdf/SIGGRAPH2001_CoursePack_08.pdf
Hi
How about this nice little learning example?
http://de.mathworks.com/matlabcentral/fileexchange/5377-learning-the-kalman-filter
Hello. You do not need a Kalman filter to estimate a parameter. You should use a parameter estimation technique like recursive least squares or one of its variants. If the parameter is in fact part of a dynamic state space, so there are other variables you need to estimate at the same time, you can use a Kalman filter. In any case, the state equation for a parameter just says that the value does not change with time, i.e. x(k+1)=x(k) and you will typically have very little process noise on such states in the KF.
it depends on the system you are working on to estimate its parameters, where Kalman filter can be very efficient for time varying parameters. have a look to this paper .
http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=6837021&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D6837021
Please read 1503.04313v1 in the arXiv and run the MATLAB programs therein.
Thank you Youssef Harkouss. Eduardo Freire Nakamura. Pavel Osinenko. Mohamed Ahmeid. Mudambi Ananthasayanam.
Thank you Graham W Pulford. I understand I do not need Kalman filter for parameter estimation. I just wanted to know if at all I uses it how will it work? Problem is solved.
Thank you all for responding.
Please see the reports arXiv:1503.04313v1.pdf, arXiv:1505.07201v1.pdf and arXiv:1505.07208v1.pdf.
and the two papers in SADHANA December 2016 issue.
this is my study in hydrology, but I think the application of FK has the same principle:
https://www.researchgate.net/publication/267326130_Multi-site_modeling_and_prediction_of_annual_and_monthly_precipitation_in_the_watershed_of_Cheliff_Algeria
Hello Dr. Rofatto!
Last year I found a interesting material about Kalman Filter, very basic to us, but very good to teach in a course based on applications.
Please, take a look here https://www.kalmanfilter.net/default.aspx
Since we've a slight modified KF related to notation, to our applications, the method is only one, an heritage of Electric Engineering.
Glad to find you here with good questions about concepts that we have to face every day : )
Kind regards
Guilherme Poleszuk
Dear Guilherme Poleszuk (alias Pole),
Thank you so much for sharing this great educational material about Kalman Filter.
I hope that you will come and visit us here at the UFU!
Best Regards,
Vinicius F Rofatto.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Control Theory
- Control Systems