EKF-Quaternion-Attitude-Estimation icon indicating copy to clipboard operation
EKF-Quaternion-Attitude-Estimation copied to clipboard

Documentation and help

Open alireza787b opened this issue 4 years ago • 9 comments

Hi great work I have already done the state estimation with linear Kalman filter... Now I want to do it with your Code (EKF) and publish the comparison... but I need more description of how you matlab function works I added that to my simulink model as matlab function that updates each iterates I have a few questions: What is Va? How should I know the velocity?? I dont have the velocity value :| I ran the simulation with : Va , MagDec = 0 NDivT =1 but I got NaN value as output

Can you please help me with an example for implementing your code in a working project? I will publish the comparison result in youtube and put the link to your github as source Thanks

alireza787b avatar May 23 '20 19:05 alireza787b

Hi! I have the same problem, how do you solve it? Thanks!

Vincy-Ying avatar Dec 16 '20 08:12 Vincy-Ying

I couldn't solve it.. I used a custom AHRS block.

alireza787b avatar Dec 16 '20 11:12 alireza787b

I couldn't solve it.. I used a custom AHRS block.

Thanks for your reply, I have the problem when using the Kalman Filter, do you know how to solve the singularity in the Euler angle? Thanks!

Vincy-Ying avatar Dec 16 '20 12:12 Vincy-Ying

use Quaternian based methods: https://github.com/alireza787b/SimulinkEKFAttitude

alireza787b avatar Dec 16 '20 13:12 alireza787b

use Quaternian based methods: https://github.com/alireza787b/SimulinkEKFAttitude

Thanks for your reply. I have already tried this project. However, in this project, although the Quaternian has been used as the state instead of Euler angle, the Euler angle transformed from the Quaternian also have the singularity, just shown in the 11 figures from the Matlab. Do I need to change the 'ExampleScript.m' ?

Vincy-Ying avatar Dec 16 '20 14:12 Vincy-Ying

I am using the method in this project (it's my own repo) and it's working perfectly even at 90 degrees roll and pitch.

alireza787b avatar Dec 16 '20 14:12 alireza787b

I am using the method in this project (it's my own repo) and it's working perfectly even at 90 degrees roll and pitch. Thanks for your repo. image When the theta at 90 degrees, the phi and psi change a lot. Do you solve it?

Vincy-Ying avatar Dec 16 '20 14:12 Vincy-Ying

I think This is not singularity... This happens because the value is limited to + or - 180 you can play with the value around the boundary.... singularity happens at +- 90 and I don't see a singularity at 90

alireza787b avatar Dec 16 '20 15:12 alireza787b

I think This is not singularity... This happens because the value is limited to + or - 180 you can play with the value around the boundary.... singularity happens at +- 90 and I don't see a singularity at 90

Thanks for your reply, I will try it again!

Vincy-Ying avatar Dec 17 '20 10:12 Vincy-Ying