Зарежда...
1 харесван файл
3 коментара
0 видеоклипа
0 качвания
0 последователи
  • Default

    @LeFix If the plan is to make the PID terms user adjustable then that would be the best option I guess. Not sure if you can model the I and D terms though (and probably not needed). Even with adjusting the strength of the P term it will be simulating the inertia by simply introducing more delay to the error response which will have the effect of making the quad feel "looser". Not exactly the goal but we'll have to see how it works out.

    27 ноември 2017
  • Default

    @LeFix No, not a deadtime. A simulation of the inertia required to get the axis rotation accelerated. In a physical system it takes time to accelerate a mass due to its inertia. The P term calculates the needed "push" to get to the desired rotation rate, but the axis doesn't immediately achieve the desired rate due to the inertia. Additionally the inertia will also cause an overshoot of the desired rate which D is trying to compensate for. So ideally the effect of the inertia on the changes in axis rotation should be accounted for.

    27 ноември 2017
  • Default

    @LeFix As a real FPV quad pilot/racer I think the physics are pretty good. But one area that could be improved a little is how the craft responds to command inputs and the targeted roll/pitch/yaw rates. In the real world the craft has inertia and takes time to accelerate to the requested rate. In the mod it appears that the craft reaches the requested rate immediately. This has the net effect of making the rates "feel" much faster and contributes to a slightly unrealistic feel. So a simple fix would be to add some acceleration/deceleration to any rate changes on each axis.

    22 ноември 2017