I’m doing my final year in aeronautical engineering
I have a project in mind(for my degree) and it involves controlling an rc with matlab/Simulink and also maybe use image processing(very basic level). The background is, this will be an individual project, I will not be working with a group and hence I have to do all the work. So I want to keep it simple and impressive.
I’m relatively new to the world of rc aircrafts and I’ve ordered my first trainer to start flying J
Things which have been bothering me and I need help on:
What Rc aircraft would be ideal for the job?
Advice on the setup required would help a great deal.
How would I imbed the codes and what platform would I have to use?
For image processing, I want to implement a very basic system just to prove that an advanced version can be done.
Using the Ardupilot line is a big advantage, but I fear that it might seem that I put in less effort since most of the functions are already programmed into it?(Is there any other way I could do this, since it is a project for my degree, I have to do some amount of work)
As I mentioned, I’m very new to this part of the woods, it would be great if you guys could point me towards the right stuff required for this feat. Your comments/advice/thoughts are welcome and appreciated.
Thanks in advance.
Well, I'd start with the ArduPilot. If you try making your own autopilot then you'll be stuck for at least a year learning all the intricacies of it.
If your going to try to do stuff via Simulink, you'll have to have a way of communicating between the craft and the PC. ArduPilot has a few things, but what your wanting to do requires a fair share of bandwidth. (I think your trying to do a hardware-in-the loop testing, yes?)
Starting off, which craft do YOU want to tackle?
I was not planning to make a full functional autopilot system, I was thinking more on the lines of giving the craft certain commands like holding altitude or following a waypoint,changing altitude etc.... Like I mentioned, it will be my final year project so I hope using Ardupilot won't seem like I haven't done much work(since it's a pre-fabricated autopilot system). Yes, I guess you could say it's a HIL simulation test.
About the crafts i'm stuck between the bixler and the easystar. Since they both are excellent with FPV systems, I guess they would be perfect for the job.
And I'm not sure how to embed the codes and control the craft using them. Any ideas?
Well, you need to first build up some sort of functional diagram or a flow chart that explains how your control systems will work. Once you have this figured out to your liking, you can then proceed to coding it with minor trouble.
How much do you know about control systems (especially discrete-time control systems), by the way?
I'm very new to this field, but I'm familiar with control systems and I will get help+mentoring with it. I'm more confused about transmitting the codes to the craft, since I'm still researching about the project are there any books/web-sites where I could obtain more knowledge about these type of systems?
I haven't come across any books that discuss it, but I would imagine the codes are proprietary to each developer. You might pick up a few ideas in the Ground Station forum here.
The codes are not the problem, it's just that I need a way for the pc to communicate with the carft. I will snoop around, but if you come across anything related, please do let me know.
Thanks for the help.
You can talk directly to the craft via Windows Hypertermal and using an Xbee transceiver.
You can make your programs treat the Xbee just like a traditional modem, add support for a COM port (a.k.a Serial Port) and there ya go.
There's paparazzi, which is much more difficult to learn but more advanced and customizable. According to many, most of the original code for ardupilot came from paparazzi.
That's not a slam on anyone, but I've heard from several people that ArduPlane started as a windows version of paparazzi backported to 8-bit processors.
As far as what aircraft to use... Are you serious? I don't think you'll survive your defense if you just slap an autopilot into some ARF kit. It's really not that hard to buy some styrofoam wings (flyingfoam.com maybe?) and carbon fiber tubes and slap them together into a plane.
If that seems too hard to you just look at the pizza-box-flyers (PBFs). People are literally flying square pieces of cardboard.
Hmm yes I guess that's true, I'll probably cad model a craft and then order the foams so it'll be much more hands on and self built.
Flying wings are a good idea don't you think?
Flying wings are the most efficient, so bonus points there. They also aren't flyable without computer control, so that's proof that the autopilot is doing something useful.
You got me there. But fly-by-wire systems and/or autopilots are the only real way to make them really viable. Pre computer control they were notoriously unstable and difficult to control. They were never safe enough to even be considered for most aviation uses.