Tilt a FMA Thermopile Sensor for Aircraft Control???
Hello,
i am just starting with my first UAV-Foam-Flyer these days.
Here is a idea i had:
The FMA Co-Pilot seems to be a popular way of stabilizing pitch and roll axis.
My idea is to use the sensor for aircraft pitch controll too.
If i tilt the sensor by (eg.) 2deg. during flight with a servo, the plane should go in a 2deg climb/descent.
In that way i could control the planes attitude with the remote control, OR a autopilot. i don`t drive the elevator itself.
If i control airspeed with a sensor (gps, air-pressure-probe) i should have a more direct control of the AC-altitude, as just with power alone.
Has anybody experiences with that??
P.S.: a tilting fma-sensor could also be used to estimate the planes bank and pitch angle.
i am just starting with my first UAV-Foam-Flyer these days.
Here is a idea i had:
The FMA Co-Pilot seems to be a popular way of stabilizing pitch and roll axis.
My idea is to use the sensor for aircraft pitch controll too.
If i tilt the sensor by (eg.) 2deg. during flight with a servo, the plane should go in a 2deg climb/descent.
In that way i could control the planes attitude with the remote control, OR a autopilot. i don`t drive the elevator itself.
If i control airspeed with a sensor (gps, air-pressure-probe) i should have a more direct control of the AC-altitude, as just with power alone.
Has anybody experiences with that??
P.S.: a tilting fma-sensor could also be used to estimate the planes bank and pitch angle.
Replies to This Discussion
-
3D RoboticsPermalink Reply by Chris Anderson on -
You mean tilt the FMA sensor with the autopilot? (You can already override the FMA co-pilot with RC, so if you want to move the elevator manually while under co-pilot control you can).
I guess I'm still a little unclear about what you're trying to achieve. Is it that you're uncomfortable using throttle for altitude control under autopilot?
-
Permalink Reply by Tom in ON on -
I think what some of us are trying to figure, if we there is a way to have the FMA co-pilot help the airplane stay stable during manouvers, for example during a 10 degree right turn. The FAM co-pilot would make sure that the turn stays between (let's say) 8-12 degrees as long as the manual control calls for a 10 degree turn. Basically instead of trying to keep (for example) a straight flight at 0 degrees turn (allowing -2/+2 degrees deviation from zero) we introduce an error that makes the FMA co-pilot think that we are trying maintain straight flight at 0 degrees while in reality we are turning at 10 degress.
This way the airplane stays under the FMA co-pilot control at all times (for example allowing -2/+2 degree deviation in any direction from the actual manual control setting), even when manually making the plane deviate from level flight by making the plane fly up or down (for example).
For example, if we want to put the plane in a 20 degree steady drive, the FMA co-pilot would keep it there thinking that it's keeping the plane in level flight. Would have to offset the error that a manually-induced deviation (i.e. turn/descent) would cause in reading the (in the horizon sensing) variables that the FMA co-pilot is using to keep the airplane level. That way the FMA co-pilot thinks it's maintainaining a level flight, while indeed it's maintaining a 10 degree turn with the same +/- 2 degrees accuracy. Hope this makes sense...
-
Permalink Reply by Longshot7601 on
-
This makes perfect sense. Make the aircraft bank left 20 degrees by tilting the sensor to the right 20 degrees relative to the airframe.
I was kind of looking to this type of system for a cheapie camera stabilzation platform. Let the FMA Co-pilot hold the platform stable in pitch and roll and have a rate gyro take care of the heading.
Has anybody done anything similar?
-
Permalink Reply by Tom in ON on
-
Longshot,
I think that's what HugePanic originally proposed as well. What I was wondering if actually making the plane tilt like 20 degrees with the sensor fixed solid on the airframe, but write the code as such that we offsett the diffrence generated by the turn in relation to the sensor signal back to "zero" so the processor thinks we are level and trying to maintain that "level: flight while in reality we are at 20 degrees. Of course when we remove the manually generated signal we wouldn't need to generate any offsett anymore and we would be at real level flight.
-
3D RoboticsPermalink Reply by Chris Anderson on -
Just to be clear, there seem to be two thread of thought here:
1) Tilting the FMA sensors to help stabilize manual flight
2) Having the autopilot tilt the sensors so you can directly control the ailerons and elevator, rather than just having two disconnected systems (FMA on ailerons and elevator; autopilot on rudder and throttle) that are sometimes at war with each other.
The FMA CoPilot already does #1, so I'll move to #2. Yes, this could work, but it's a bit of a kludge. ArduPilot Pro is designed to achieve the same thing but without the complicated mechanical tiling gimbal that would be required. But, in fairness, we haven't released ArduPilot Pro code yet, so perhaps a kludge is better than nothing ;-)
Just out of curiosity, what's the problem you're trying to solve? The somewhat graceless and slow "crab" turns that occur when an autopilot-controlled rudder fights the FMA-controlled ailerons?
-
Permalink Reply by HugePanic on -
I still haven't "seen" any homebuild UAV until today.
But i like the idea (i hope i would work that way) that i have more data available.
If i bank that sensor by 20° (by remote or by the autopilot) i know that the plane will bank 20° +- 2°. I don't have to apply "some" rudder/aileron and "see" what happens. i even don't need a controler that compares target and actual angles.
the funktion "servo-signal to bank angle" can be work out at the ground, and stored in a table in the code. So i can command a 20° bank angle directly.
In my original idea i had in mind to use it for climd/dive maneuvers.
It should be easier to climb/descent to a target height, when you can command an angle.
-
Permalink Reply by Marty on -
Boomslang (another memeber on here) started looking into this a while ago. He provided this link: http://www.uavs.net/horizon_sensing_autopilot.pdf
Which is a paper some guys in Melbourne wrote on the subject.
-
Permalink Reply by Scott James on -
I had a similar Idea here. I never got the Basic Stamp code working though.
-
3D RoboticsPermalink Reply by Chris Anderson on -
That looks right. Two thoughts:
1) If you're going to use ArduPilot, the switchboard is built-in
2) Have you thought about what kind of mechanical gimbal you want for the FMS sensor-tilter? Would a pan-tilt camera mount work?
-
Permalink Reply by HugePanic on
-
1) I just have ordered a Arduino 1000er board. So i start my work with that. Thanks for the tip with the arduPilot, maybe i'll switch later.
The switchboard is not necessary for the first tests.
2) I will tilt the sensor only on one axis first. That makes it easier.
I don't know if i need two axis movement. It seems that a lot of UAVs turn with rudder only and wings level. this is my first attempt too.
One more question:
I just finished the wing. I installed only one aileron to keep it all lightweight and simple. I hope not to fly many stalls, so i hope this is ok.
What do you guys think about this?
-
3D RoboticsPermalink Reply by Chris Anderson on -
Just doing tilt seems like a reasonable proof of concept. But I would add that other aileron ;-)
Pls upload pictures of the project in process. I'm sure others would like to follow along.
Social Networking
Top Content
Contests
Season Two of the Trust Time Trial (T3) Contest has now begun. The fourth round is an accuracy round for multicopters, which requires contestants to fly a cube. The deadline is April 14th.A list of all T3 contests is here.
Groups
-
Moderators
24 members
-
ArduCopter User Group
1355 members
-
ArduRover User Group
207 members
-
Documentation Editors
52 members
-
ArduPlane User Group
719 members
Advertisement
© 2013 Created by Chris Anderson.
Powered by
