AP_Mount: Siyi gets rate map, PI control, accel limit

[rmackay9]

This improves the Siyi gimbal control inspired by the work done in the MAVProxy Siyi module:

• Rate mapping table compensates for the non-linear rate response of the gimbal
• Angle controller is enhanced in these ways
  • P controller is extended to a PI controller to reduce long term error
  • Rate targets are acceleration limited to reduce impossibly fast changes and thus reduce overshoot
  • I-term build up is avoided by using limit flags
• Debug logging of angle and rate has been added to somewhat ease support if this doesn't go well. This debugging is off by default

This has been tested on a Siyi A8 and some screen shots are shown below

Desired vs Actual Pitch Angle

Desired vs Actual Yaw Angle

Desired vs Actual Pitch Rate

Desired vs Actual Yaw Rate

[rmackay9]

Hi @timtuxworth,

I wonder if you'd be able to help me test this? I've bench tested with an A8 and I can bench test with a Z10 as well but a flight test is slightly more difficult for me.

No pressure of course but if you can help, just ping me with your autopilot and I'll create a binary. Discord also works of course.

[timtuxworth]

Hi Randy - I currently have the A8 Mini installed on an EDU-650 with a Cube Red. I'd likely be able to test it on the weekend.

[rmackay9]

Hi @timtuxworth,

Thanks very much!

Here is a binary for the CubeRedPrimary based on this branch. https://www.dropbox.com/scl/fi/k7f1yxtqa7y50ebtdu256/copter-460-siyiPI-cuberedprimary-27Jun2024.zip?rlkey=ae1mfjr1ybq3n5njpcad43lug&dl=0

I've rebased this branch on master as well.

[timtuxworth]

I tried testing this but found that this firmware seems to break NET_Px_PROTOCOL access to the A8 Mini. I have my gimbal connected to the FC via ethernet. The symptom is that it reports "Mount: not Healthy"

I proved this by flashing back to 4.5.0 beta 3 which worked.
I just put 4.5.4 stable on, which also works, the FC can communicate with the Gimbal using stable, but with this PR it reports "Mount: not Healthy".

[rmackay9]

Hi @timtuxworth,

Now that we've resolved the networking issue, I wonder if you'd be willing to test this change again? I've put a new CubeRedPrimary binary here which is this PR rebased on master
https://www.dropbox.com/scl/fi/l0apmxdhfom4ksen95rx3/copter-460-siyiPI-cuberedprimary-17Jul2024.zip?rlkey=1xajqqvzd0wa6njyajcuki9nk&dl=0