Subject:
|
Balancing Bots
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Tue, 4 Jan 2000 00:00:18 GMT
|
Viewed:
|
819 times
|
| |
|
|
I'm working on a 'bot that will balance itself, and having pretty good
success (it will keep itself up for 30 seconds or so on a good run,
traveling around a bit, before it falls over). But I'd like to make it
more stable, if possible, so that I could nudge it and it would correct.
Basically, it's a long (about 2 feet) beam construction, with the RCX on
the top and 2 wheels on one axle at the bottom. I'm using a homebrew
angle sensor (with an "outrigger") to detect the angle it makes with the
floor, then a motor driving the wheels corrects to get it vertical again.
I have 2 questions:
First, has anyone done this king of thing before? The mechanics are
relatively simple, and I think I have a pretty good handle on it, but the
programming is a challenge: I'm walking a thin line between it being too
reactive (vibrating constantly) and not reactive enough (if it waits
until the off-angle is too large before correcting, it can't "catch up").
Any ideas?
Second, To make the reaction speedy, I basically have to gear up the
motor, but power suffers (and correcting from a large off-angle requires
power AND speed). Has anyone ever tried driving the same axle with 2
motors simultaneously? That seems like it would be ideal, but I know the
motors are never identical and having them "fighting"each other would be
a bad idea. Maybe some sort of differential arrangement...
Thanks,
Dave Johnson
|
|
Message has 5 Replies:
 | | Re: Balancing Bots
|
| From: "Dave Johnson" <djohnson@sirius.com> (...) You can easily do this with a differential. Use it the opposite way of the way that it is used in car models. Hook each motor up to the side of the differential, and then run your wheel off of the (...) (25 years ago, 4-Jan-00, to lugnet.robotics)
| | | Re: Balancing Bots
|
| Dave Johnson wrote in message <0FNS00E81AOHFT@mta1...bi.net>... (...) Nice!! Sounds like a special wheelchair Bob Fay showed me a URL for (which I don't have handy, unfortunatley...). It was a really clever chair. It could stand up on two wheels so (...) (25 years ago, 4-Jan-00, to lugnet.robotics)
| | | Re: Balancing Bots
|
| Dave Johnson <djohnson@sirius.com> wrote: [...big snip...] (...) You can safely drive the same axle with more than one motor. Yes they are never identical but this won't cause you any problem. They simply will have a (slightly) different load, but (...) (25 years ago, 4-Jan-00, to lugnet.robotics)
| | | Re: Balancing Bots
|
| (...) Take a look on (URL) Krister -- You have moved your mouse. Windows must be rebooted for the changes to take effect. (25 years ago, 4-Jan-00, to lugnet.robotics)
| | | Re: Balancing Bots
|
| (...) It's an inverted pendulum - right? (...) I made such a project at university, but it didn't involve LEGO. I was working with a group and we tried different approaches. Our best solution was a Fuzzy-PD controller. The differential angle part is (...) (25 years ago, 4-Jan-00, to lugnet.robotics)
|
10 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
