Modify coefficients for better output of AHRS in "UAV Toolbox Supportage Package for PX4 Autopilots", in Simulink

3 views (last 30 days)
Hi everyone;
My teamwork and I are working on the Holybro "Pixhawk 4" board, with its add-on "UAV Toolbox Supportage Package for PX4 Autopilots".
We would like to use the predisposed blocks of this add-on to estimate the attitude of the board (and consequently of the drone) and by testing the various tools provided we have identified "PX4 Vehicle Attitude" as the best block, reading the uORB topic "vehicle_odometry".
We know that the output of "Vehicle Attitude" estimates the attitude using accelerometer, gyroscope and magnetometer producing a quaternion.
The values are good but we need more responsive dynamic parameters.
How can we change the filter coefficients implemented by the block?
Thanks

Answers (1)

Arun Mathamkode
Arun Mathamkode on 15 Jun 2022
As you mentioned correctly, the block just reads the attitude values from the uorb topic 'vehicle_odomtery'. The block doesnot implement any filters and hence if you want to modify the filter values probbaly you may need to investigate which px4 module is publishing the topic 'vehicle_odometry' and you will find the filter implementation in that specific module. To address these kind of usecase, PX4 have provided a set of 'PX4 Parameters'. To start with you can have a look at the Paramters for Attitude Q estimator. You can update the parameters through QGroundControl or the nsh console.
  1 Comment
BERNARDO INCERPI
BERNARDO INCERPI on 15 Jun 2022
Thank you for your reply.
We have difficulty figuring out which uORB to use since no one seems to publish values with the filters turned on.
We know, through QGroundControl, that you can turn filters on or off and adjust them but we do not understand how to do this in Simulink. It is not clear to us how to pass the parameters found in QGroundControl inside Simulink.
Similarly, we don't know how to investigate to figure out which PX4 module is publishing the "vehicle_odometry" topic. We only have access to the "msg" describing the block outputs, and none show the presence of the filters.

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!