EKF Dynamics Update

[Pedro-Roque]

Solved Problem

Attempt at fixing the EKF for microgravity dynamics.

Solution

For now purely deleting old solution, but can be made to accommodate in parallel.

Test coverage

• Unit/integration test: none yet
• Simulation/hardware testing logs: none yet

[Pedro-Roque]

For now, I chose to ignore merging-compatible modifications with PX4 main tree - but it could easily be done in this case. Just checking if this is the direction in which we want to go.

Still requires updates on the EKF state (we don't need or want wind estimates, for instance)

[Jaeyoung-Lim]

For now, I chose to ignore merging-compatible modifications with PX4 main tree - but it could easily be done in this case. Just checking if this is the direction in which we want to go.

Probably a copy-paste on a new module?

Still requires updates on the EKF state (we don't need or want wind estimates, for instance)

This is configurable in upstream by PX4#22172, just FYI

[Jaeyoung-Lim]

Do you think it would be better to remove the position states completely from the estimates? or do we want this configurable?

[Pedro-Roque]

I just removed the gravity component. But if we make it configurable then it can be user selected with a parameter on running the script.

[Pedro-Roque]

For now, I chose to ignore merging-compatible modifications with PX4 main tree - but it could easily be done in this case. Just checking if this is the direction in which we want to go.

Probably a copy-paste on a new module?

Still requires updates on the EKF state (we don't need or want wind estimates, for instance)

This is configurable in upstream by PX4#22172, just FYI

Then perhaps we could even make a configuration to which dynamics we should use. It would make it more functional perhaps.