OpenServo.com

[stefanengelke]

Hi all,

I have finished the documentation of the mathematical model I'm working with:
http://www.openservo.com/moin.cgi/MathModel
The model is pretty much the same Zed is explaining in his document. I have also implemented it in Matlab, but I'm trying to port it to Octave now. My objective is to write some scripts for an easy identification of the model parameters.

If we have a identified model, my next objective will be the implementation of a new control strategy. My plan is as follows:

- Using an exact input-output linearization to compensate the friction
- Introducing the external torque caused by a load as a new state in the state-space-model
- Using a state estimator (=observer) to reconstruct servo velocity and external torque
- Using a state regulator (=state space controler) as control-loop

Yes I know, this is a comprehensive plan and will take some time to be implemented, but this strategy seems very promising for me.

Stefan

[ginge]

Hey stefan, good to see you back!

The wiki entry is amazing. Very concise!

Thanks for this information, it will help me model my robot arm with some more accuracy.

Barry

[mpthompson]

Hi Stefan, I also welcome you back. It's been a little quiet around here lately, but hopefully we'll get things going again.

-Mike

[stefanengelke]

I've fixed a small mistake in the mathematical model. The friction moment depends on the sum of all other moments and not only on the motor torque.