Refactoring on pid iterm

[shota3527]

Some pid refactoring mainly addresses the accumulation of iterm. I believe this will improve flight performance on low-authority multirotor, such as vtol platform.

Changes:

1. Stronger iterm_windup, enhances the effectiveness when the value is set below the default value(50), no accumulation occurs when iterm_windup+50 of motor correction range is reached. When value is set to equal or above 50, remain the same behaviors as the current master.
2. ~~combine itermLimitActive with itermFreezeActive to itermFreezeActive. Because the actual itermLimitActive implementation has the same result as itermFreezeActive~~
3. remove fw_iterm_throw_limit and introduce pid_iterm_limit_percent
4. pid_iterm_limit_percent limit the accumulated iterm by percentage of pidsum_limit or pidsum_limit. This iterm limit is also applied to MR pid controller. The default value is set to 33%, which behaves the same as fw_iterm_throw_limit default 165.
5. add option to enable tpa on yaw axis
6. I term back caculation anti-windup can not push the I iterm to opposite direction

Have done some tests with SITL in the real flight simulator, The flight performance was terrible but become acceptable with vtol platforms

[shota3527]

we need something equivalent to IMAX

[DzikuVx]

Ah, I thought it's for MR and wanted to test it today :)

[shota3527]

after some study, i have found this pr https://github.com/iNavFlight/inav/pull/6387

thinking about revert this

combine itermLimitActive with itermFreezeActive to itermFreezeActive. Because the actual itermLimitActive implementation has the same result as itermFreezeActive

[sensei-hacker]

@shota3527 I see this was opened, then closed, then re-opened. Is it still current?