Control system implementation

For the discussion of the sciences. Physics problems, chemistry equations, biology weirdness, it all goes here.

Moderators: gmalivuk, Moderators General, Prelates

User avatar
nehpest
Posts: 518
Joined: Fri Jun 12, 2009 9:25 pm UTC

Control system implementation

Postby nehpest » Mon Nov 28, 2011 9:29 pm UTC

As part of an assignment for a club at my school, I need to design and implement a control system to keep a directional antenna aimed at a distant moving source. Since I've never taken a controls course, I'm self-studying out of Dorf's Modern Control Systems, and reading up on wikipedia as a support. Now, my problem is that even if I "design" a correct system, I have no idea how to implement it.

The actual details of the task:
We have a high gain Yagi antenna that is to be mounted on a rotating disk. The Yagi must be kept trained on an RC plane roughly 10-25 miles away, which is traversing a circular track around the antenna at speeds between 100 and 500 mph (ideally, the system should be able to track a target moving upwards of 800 mph, but that won't be important for 6-12 months). Some back-of-the-napkin calculations indicate that I should expect angular velocities between .5 mrad/s and 2.75 mrad/s. Also, our Yagi has a half-power beamwidth of about +/-15°, which gives me an upper bound on how off-course the system can be.

I have the following inputs available: GPS coordinates of the antenna (assumed to be fixed), GPS coordinates of the plane, kinematic data on the plane (heading, velocity, possibly acceleration), the amplitude of the received signal, and the gain envelope diagram from the manufacturer (the datasheet of the antenna we ordered is here).

Some questions to get this started:
  • I'm familiar with MATLAB and Simulink, and was thinking of using them to design the system. Once I have a simulink file, how can I translate it into either code for an FPGA or hardware?
  • What kind of pitfalls should I watch out for?
  • Are there better tools available than MATLAB/Simulink? What do professionals use?
  • Is this system simple enough to lend itself to my limited controls knowledge?

Discuss, and help!
Kewangji wrote:Someone told me I need to stop being so arrogant. Like I'd care about their plebeian opinions.

blag

User avatar
jmorgan3
Posts: 710
Joined: Sat Jan 26, 2008 12:22 am UTC
Location: Pasadena, CA

Re: Control system implementation

Postby jmorgan3 » Mon Nov 28, 2011 11:14 pm UTC

That sounds really slow. You could probably get away with just position control on a stepper motor, updating every few seconds. No need to design a control system.
This signature is Y2K compliant.
Last updated 6/29/108

User avatar
nehpest
Posts: 518
Joined: Fri Jun 12, 2009 9:25 pm UTC

Re: Control system implementation

Postby nehpest » Tue Nov 29, 2011 12:51 am UTC

I originally suspected that would be the case, but I was told we should build the system anyway. This definitely won't be the last control we'll have to build, so maybe we can open the thread to a more general discussion of control systems implementation. Also, I'm still very interested in the community's thoughts on those questions.
Kewangji wrote:Someone told me I need to stop being so arrogant. Like I'd care about their plebeian opinions.

blag

User avatar
The Geoff
Posts: 144
Joined: Wed Jun 08, 2011 6:22 am UTC

Re: Control system implementation

Postby The Geoff » Tue Nov 29, 2011 1:17 am UTC

My quick and dirty solution would be to take the two GPS coordinates, use them to calculate the angle, and feed the +/- correction needed to a stepper motor. I'd probably use a CD drive to do it because you've got most of the hardware interfaces in place already.

Disclaimer: I've never done anything like this, I've no idea if it would work. Just throwing an idea in there! :)

User avatar
scarecrovv
It's pronounced 'double u'
Posts: 674
Joined: Wed Jul 30, 2008 4:09 pm UTC
Location: California

Re: Control system implementation

Postby scarecrovv » Tue Nov 29, 2011 1:42 am UTC

The easy way would be iron sights and a freshman. That would probably be good enough. If your freshman is having trouble seeing the plane you could replace the iron sights with binoculars. Maybe add an audio alert if signal strength starts falling off. If it must be automated though, jmorgan3 and The Geoff have the right idea.

User avatar
The Geoff
Posts: 144
Joined: Wed Jun 08, 2011 6:22 am UTC

Re: Control system implementation

Postby The Geoff » Tue Nov 29, 2011 2:02 am UTC

More generally, this seems like a good place to start, assuming you're handy with a soldering iron and know your way around a linux box: http://linuxgazette.net/179/radcliffe.html

There's lots of cheap, off-the-peg microprocessor systems you can use to build self-contained systems. Generally you write a C type program, compile a binary, and then burn it to the chip. A popular one is called Propeller: http://www.parallax.com/propeller/

Kugala
Posts: 6
Joined: Thu May 06, 2010 6:55 pm UTC

Re: Control system implementation

Postby Kugala » Tue Nov 29, 2011 2:07 am UTC

Without knowing your update rates for the data, the typical implementation for this would be a Kalman filter. Specifically, use the velocity data to predict how fast to track your antenna, use the GPS data to apply corrections to your antenna direction, and use the amount of correction to adjust your tracking velocity. Assuming the RC aircraft isn't rapidly maneuvering during between data updates this is pretty accurate.

User avatar
nehpest
Posts: 518
Joined: Fri Jun 12, 2009 9:25 pm UTC

Re: Control system implementation

Postby nehpest » Tue Nov 29, 2011 5:20 am UTC

Thanks all for the thoughts!

scarecrovv: aiming it manually was our first thought, but a ~3 ft target at 10 miles is going to be almost impossible to see. I love the audio alarm idea - I'll probably incorporate it into the design.

Geoff: simply using the GPS coordinates and interpolating was my first stab at a technical solution, too. When we were originally talking about the problem, I made an order-of-magnitude error in my napkin math, and thought we'd need something spinning around 6 rpm, which is why we dropped that plan. With the revised estimates, it would totally work, but I'd love to make a mini project out of this and overengineer it a bit for learning purposes. Also, thanks for those links!

Kugala: thanks for the phrase "Kalman filter". It sounds vaguely familiar, but I wouldn't have remembered it on my own. We won't be maneuvering much at all; we're much more concerned with speed on this one.
Kewangji wrote:Someone told me I need to stop being so arrogant. Like I'd care about their plebeian opinions.

blag

User avatar
The Geoff
Posts: 144
Joined: Wed Jun 08, 2011 6:22 am UTC

Re: Control system implementation

Postby The Geoff » Tue Nov 29, 2011 9:28 pm UTC

I've just had a "hang on a minute...." moment.

Hang on a minute....you're planning on flying an RC plane in a circle at trans-sonic speeds?! That's really cool, what the hell is the project?

User avatar
nehpest
Posts: 518
Joined: Fri Jun 12, 2009 9:25 pm UTC

Re: Control system implementation

Postby nehpest » Tue Nov 29, 2011 10:27 pm UTC

:mrgreen:

Precisely what you said: we want to be the first civilian group to exceed Mach 1 in a UAV. In the process, we get to test an as-yet unreleased microturbine that spits out a couple hundred pounds of thrust, the JetCat P-300 (which, alas, isn't up on their site yet). If you're interested, there is a somewhat out-of-date summary of the project here. Ostensibly we want to get into the Guinness and FAI record books, but really it's just an awesome project that will make us hireable after graduation.
Kewangji wrote:Someone told me I need to stop being so arrogant. Like I'd care about their plebeian opinions.

blag

User avatar
The Geoff
Posts: 144
Joined: Wed Jun 08, 2011 6:22 am UTC

Re: Control system implementation

Postby The Geoff » Wed Nov 30, 2011 12:01 am UTC

Truly awesome, the very best of luck with it :D

Have you got a website up so we can keep track of your progress? (And if not, I can supply you with hosting on a bells 'n whistles Linux server for free if you want, 'cos I'm that impressed!)

User avatar
nehpest
Posts: 518
Joined: Fri Jun 12, 2009 9:25 pm UTC

Re: Control system implementation

Postby nehpest » Wed Nov 30, 2011 4:42 am UTC

The Geoff wrote:Truly awesome, the very best of luck with it :D

Have you got a website up so we can keep track of your progress? (And if not, I can supply you with hosting on a bells 'n whistles Linux server for free if you want, 'cos I'm that impressed!)

At the moment we're using a not-altogether-adequate Google Site. I'll bring it up at our weekly meeting tomorrow, and let you know what The Powers decide!

Edit: the meeting is Thursday, not tomorrow. Sorry, I've been a day ahead all week.
Kewangji wrote:Someone told me I need to stop being so arrogant. Like I'd care about their plebeian opinions.

blag


Return to “Science”

Who is online

Users browsing this forum: No registered users and 18 guests