Society of Robots - Robot Forum

Electronics => Electronics => Topic started by: TrickyNekro on October 28, 2008, 05:03:01 PM

Title: Building a Radar
Post by: TrickyNekro on October 28, 2008, 05:03:01 PM
Hello guys,

I know that building a radar is one of the hardest things to do, but this year I'll be building or at least try, a UAV helicopter...

I know some basics about the radars, so I'm ok with that... Thing is that helicopters usually fly low... Flying low means many obstacles...
But the hardest thing to detect is a tree... I want to give this ability to my chopper... detect trees...
Camera is not an option... In order to detect trees with a camera you must come in heavy with a nice resolution camera and such
But big resolution requires a lot of processing power... so.. I don't want to thing about it... I'll be going in micros this time...

I think of sonar... instead of a radar... but anyway...

My point is how to build a radar... Small range one... 0.5m to 10m or more is good... But I want to detect trees...

This is a big big problem... I can not use narrow beams, scanning lasers are too bulky but also cost a lot!!!, plus I'm going outside sun is a factor...
So I can only depend on a radar system (which has a fairly big range at least...) or a sonar that will give me at least a warning...

That's it, any ideals are more than welcomed....

Thanks in advance,
Lefteris, Greece
Title: Re: Building a Radar
Post by: Rebelgium on October 28, 2008, 05:25:23 PM
A reasonably good sonar detector can detect fine objects like ballpens, so I believe it'll do good detecting trees.
I think the simplest way is to mount three of four sonar sensors next to eachother on the front of your helicopter. Considering each sonar sensor has a nice amount of ° range, you should get a good "view" of your surroundings.
Title: Re: Building a Radar
Post by: ArcMan on October 28, 2008, 08:32:06 PM
I don't know if technology has changed alot lately or not, but I remember working with a radar engineer about 20 years ago when I was an intern.  He had some very specialized stuff - horns and waveguides - crazy complicated.
Title: Re: Building a Radar
Post by: dunk on October 29, 2008, 04:51:46 AM
the problem with radar for this sort of application is you RF waves move at the speed of light.
because of this you will need some crazy fast electronics for close range work.

the problem i would see with sonar is your UAV is a noisy platform. i'm not convinced sonar would work over the noise/vibration.


dunk.
(sorry. no helpful suggestions. just negative comments.)
Title: Re: Building a Radar
Post by: gamefreak on October 29, 2008, 07:14:19 AM
Also, when moving you would get distorted results, if you send a sonar bast and then move then the one you are getting may be too late to warn you, not to mention that sonar is echo(y). With many obstacles you could be recieving many wrong readings. Then again all of this paragraph is based on what i'm imagining, and I have been wrong before...

What about something like Lidar? Police use it out in the sun and it works fine.
Title: Re: Building a Radar
Post by: TrickyNekro on October 29, 2008, 11:28:40 AM
Hello again...
First of all, thanks for any kind of comments...

Lidar is an option... thing is that I want a lightweight solution... lidars are a little bit bulky.. as I said before bulky thing is not an option...
we talk about a heli that is made for 3D flights but can not lift that much...

Unfortunately... sonar is not a good solution too... although this one will detect trees, movement is a factor...
This is a hard one... I must know the speed of the helicopter anytime... so I can measure distance and speed making a nice calculation for
impact time... This is tough... very though... but manageable...

At this point I have a very nice idea to calculate velocity... But I need to know something...
Can I convert frequency to voltage??? If yes, I'll start a new topic to explain...
Title: Re: Building a Radar
Post by: arixrobotics on October 29, 2008, 05:44:27 PM
Well, when the normal engineering approach is limited, we turn to BIOLOGY for the solution!

Check this out http://www.centeye.com/index.html

And don't forget to look at their videos too. They have exactly what you want.

Biomimetic systems is the future!

PS: I am working on a project which is related to biomimetic vision systems at uni. Something like what Centeye did, but we are not that advanced yet.
Title: Re: Building a Radar
Post by: hudbrog on October 30, 2008, 05:18:41 AM
Can I convert frequency to voltage??? If yes, I'll start a new topic to explain...

Sure you can. Microchips AN795 as example.
Title: Re: Building a Radar
Post by: TrickyNekro on October 30, 2008, 12:35:46 PM
Biomimetic systems is the future!

So you suggest using optic flow... That's not bad... but it's sure tough... I have never used such systems before...
And perhaps I should try to experiment, a with take time... But I bet it's worth it...
You bet man... although I fear range is very short, I'll try giving it a shot!!!


Thanks again for your replies,

And if you have something to suggest please feel free to say it...