search

PX4 / PX4-user_guide

PX4 User Guide
https://docs.px4.io/main/en/index.html
Other
324 stars 1.69k forks source link

Detailed explanation of the data fusion / Kalman / attitude and speed filter algorithms as a series of notebooks or webpages? #1797

Closed jerabaul29 closed 2 years ago

jerabaul29 commented 2 years ago

First, many thanks for an awesome project.

I was wondering if it would be possible to also integrate an 'education / vulgarisation/ pedagogical' part to this project, by describing in details the algorithm, dynamic models, physical hypothesis, and implementation used for the attitude / speed Kalman filter implemented by PX4.

I think that there are some really excellent materials to learn Kalman filters, such as https://github.com/rlabbe/Kalman-and-Bayesian-Filters-in-Python , and for learning AHRS, such as https://ahrs.readthedocs.io/en/latest/filters/ekf.html , but these do not go as far as the filters implemented by PX4.

I think that having similar materials focusing on explaining / showing the derivation, underlying hypothesis and physics / math, and documenting how exactly the PX4 filters work, would be very useful for 1) people who want to learn about the topic, 2) helping new devs / contributors get up to speed.

Any way something like this could be provided? :)

hamishwillee commented 2 years ago

Thanks for posting. I agree that it might be useful, but it is unlikely to be provided unless you or someone like you does so. I'm therefore going to close this.

@priseborough Is there anything you might like to add about EFK2 learning materials you would recommend. We could add them at the end of the doc near your further information section: https://docs.px4.io/master/en/advanced_config/tuning_the_ecl_ekf.html#further-information