Closed Bengt-M closed 8 years ago
@Bengt-M I am happy that someone got excited and continued the unit tests work. So not stepping at all. If you allow me I would do some suggestions nevertheless.
Regarding the non tricopter_mixer unit tests you fixed, they started to fail in betaflight. You could consider checking if they still are and maybe create a PR to fix them in the origin as well. Just an idea.
And go ahead on the unit tests. I still need to understand the meaning of the tricopter_mixer awesome equations before writing tests (otherwise have no idea what I am testing).
@lkaino do you have any graphical representation of how you deduced the servo model and the torque estimator/compensation? How did you reached the equations used in the mixer?
Awesome! Thanks guys, I will merge this once I have tested the cleanflight master "merge".
@pqueiros: unfortunately now drawings exist of the equations. I have written a short description here: http://rcexplorer.se/forums/topic/debugging-the-tricopter-mini-racer/page/18/#post-19865.
I think this one is outdated and can be removed. There are still issues with unit test but this is not the way to solve them.
I hope I'm not stepping on your toes @pqueiros but when I started I just couldn't stop. This is a continuation of your unit_test_project branch. Now the whole suite will run. Some of the test cases are nothing more than placeholders but some actually test something. Note some strange math errors. Haven't checked if there are real errors. /Bengt