I had a full day of autonomous flight testing in twenty knot
winds, and it's clear the default PID gains in ArduPilot were not
superior ample to deal with any significant wind. The current gains
are as follows:
Please note that with this particular higher rudder get http://www.shopbestshoes.fr/Nike-Zoom-Lebron-James-9-c3744.html,
it truly is vital to ensure that your FMA Co-Pilot is performing
its occupation with wing-leveling. As soon as you are within the
air, set it on the highest get it may consider with no oscillating
(if it is a clear day and you have calibrated it ideal, that ought
to be the greatest rotation of one's get knob on your
transmitter).
A different matter with employing ArduPilot one.0 in higher wind is
always that altitude hold is pretty sloppy and it's trouble
building progress upwind. That's since ArduPilot 1.0 just utilizes
the throttle for altitude control. Once the plane is going in to
the wind, it tends to rise. ArduPilot will then lower the throttle,
but in a superior headwind, the plane may perhaps not descend. As a
substitute, it is going to just hover or perhaps move backwards
somewhat.
That's why we're shifting to elevator+throttle altitude handle
while in the ArduPilot 2.0 program. With the plane will point its
nose down a tiny bit and continue to keep the throttle likely to
make headway against the wind. It's a a lot better way to control
altitude. Jordi's testing it this weekend and may have the code
prepared to release this week.
3D Robotics Comment by Chris Anderson on March 9, 2009 at
10:08am
Imagine the PID Max and Min numbers as volume controls. The
navigation algorithm will move the rudder proportionally (inside
the Max and Min variety) towards the course the plane really should
fly. The wider the Min Max range, the "louder" the signal as well
as increased the rudder throw. Those constants that you just can
modify in the initial tab are utilized in the PID tab in lines like
this:
heading_output =
constrain(heading_output,heading_min,heading_max);//limiting the
output.
I would suggest the values must be between 15 and 30, depending on
the aircraft and conditions. The reduce the amount the additional
gentle the steering, which can be superior if you do not have FMA
stabalization or for those who stabilization isn't doing work
really well. But if it truly is as well gentle, it'll be conquer
because of the wind, as was the case together with the setting of
15 on our Superstar. Having said that 45 can be as well a great
deal and could cause spins and various instances of stabilization
failure.
And also to response your last query, this setting is indeed
analogous for the adjustable get pots on other autopilots.
Comment by smguest on March 9, 2009 at 6:37pm
I mentioned some initialization troubles and wasn't confident if it
had been due to the fact I was applying the GPS from my OSD (plus
the OSD wanted an extra sentence the ArduPilot did not),
nevertheless it seems that I was overwriting the parser's "buffer"
variable till I declared it to get 128 characters (char
buffer[128]).
I am not confident that it was essential but I wound up including
the following code following the yaw_control(http://www.shopbestshoes.fr/Nike-Zoom-Lebron-James-10-c3812.html
) statement while in the init_startup_parameters() method to be
certain I was receiving dependable readings before I set the
startup position variables.

