On 20/02/13 13:03, Jon Stephan wrote:

Ah, I hadn't thought about using stepper motors (I've never looked into how difficult the control would be). Would this package be of any use to you? I would assume you could kludge it to work, but I don't know if it really helps.

Without encoders it would be a nightmare to simulate their value from the odometry we generate...
I don't envision rewriting the low-level code anytime soon.

I will have to think about the holonomic controller. (I don't have access to that paper here, I'll get it later today) That will be difficult to debug because I don't have a holonomic robot. But I suppose I could use the holonomic controller on a non-holonomic robot, right?

Holonomic platforms are not very common [1][2].
We don't have one. That'sjust why it may be convenient to simulate one.
Holonomic platforms are widely used in theoretical works to simulate functions and theorems, thus to back your simulation with experimental results, you need to have your control law command real (non-holonomic) robots.
Hence a holonomic control for nonholonomic robots as I wrote it in my previous email.

BR

[1] http://www.oemoffhighway.com/article/10632736/movement-without-limits
[2] https://encrypted.google.com/search?q=wowwee+tribot&tbm=isch

Eng. Claudio Carbone
Embedded Systems Design

P.IVA: 11688471009
tel: +393809017424
email: Send email

My Portfolio