*Ideally* I would like to build an sudo-autopilot for a RC plane. I basically want to create a program so I can use my laptop (hooked up to the radio) as my controller. My goal would be to get it to the point where i can type "take-off" and the plane would take off, or "land" and the plane would land. Obviously there will be MANY restrictions to where and when this would work on an RC plane since clearly I would have no radar or advanced control systems on the plane. For instance the place I'm flying would have to be completely flat land since it will not be able to detect where the ground is, and I'm not looking for it to be able to handle a ton of wind.

So what I'm wondering is if there are any suggestions on what to use for positioning. GPS I know is a definite possiblilty but isn't it only accurate to with in 20 feet or so. And how well does it show height change since that is the direction I need it to be most accurate in, I figure I can *zero* ground level before it takes off. Any comments on this or any suggestions on other means for positioning data? Thanks

__________________
This post was made using 100% recycled electrons

 

The estimated position error of standard GPS in altitude is about 2X its lat/lon EPE. Differenctial gps would be more accurate, as would a regular electronic altimeter, which is probably the easiest option.

 

Yeah, I'm just trying to keep the weight down (for obvious reasons) so I was wondering if I could get away with having something for both position and altitude, but I wouldn't be suprised if in the end I have to put an altimeter on it as well.

__________________
This post was made using 100% recycled electrons

 

It would help if you tell us a little, or rather a lot more about the airframe that you are planning to use. Small RC aircraft tend to be very unstable in flight. After all , it's no good just telling the rudder servo to turn according to the flight plan if the plane is nose down at at full throttle !Accelerometers have come down in both price and size these can provide the raw data required for a navigation system. The downside is that the array must be gyro stabilised in order to provide useful information, the "head" from a scrap VCR makes a good flywheel and often has a flat/pancake motor built in depending on make/model.

 

Currently I don't really have a layed out plan. I was just trying to get an idea to see how feasible it is. I am considering doing this for a senior design project for Aerospace Engineering and plus because i think it would be very interesting. I definately would use one of the largest top-wing trainers with a large engine. Specific size or model I haven't looked into yet. Probably something along the lines of a PT 60 or so.

__________________
This post was made using 100% recycled electrons

 

I'm a RC pilot myself I think you'll need some sort of combo between radio control and computer control ie the servo's are connected to the onboard computer and to a RC receiver.

Do a search on "Chris Good; UAV project". He has done a flying wing with ducted fans and a 3 metre wing span. So far it is the best implementation I've seen.

__________________
Stef

 

OK found the site; see he has updated it as well...

http://www.chrisgood.com/uav/

__________________
Stef

 

Worth a look , but be warned it's very expensive
but it does show what can be done
http://www.micropilot.com/

 

OK OK micropilot is now officially the best implementation in my book 8) 8)

For DIY purposes, check the site I've listed above; it has formulas etc 8)

__________________
Stef