Hi every body,
I have a nonlinear system model in simulink. This system is unstable and I want to design a controller to stabilize it. Due to this, I want to linearize this model and then design a stabilizer controller. To do this, I use the linearization ability of simulnk, but Is linear analysis in controller design (in MATLAB simulnk) applicable to unstable systems ?
As far as I know, you can. I've used it for designing controller for unstable plant.
Regards,
Davood, sure you can control unstable systems. In fact, a lot of real examples of systems are unstable like, say, rockets. It does not stop us from effectively controlling them. You can analyze your closed-loop system and place poles as you want to by adjusting the feedback gain. Please read here: http://de.mathworks.com/help/control/ref/place.html.
Also, for a nonlinear system I'd first try exact linearization. This may be quite helpful. Read here: http://en.wikipedia.org/wiki/Feedback_linearization.
http://en.wikipedia.org/wiki/Feedback_linearization
http://de.mathworks.com/help/control/ref/place.html
Thank you Amin,
Thank you Pavel,
I linearize my system. It lead to a 2*2 unstable model with 8 states. The model is high order. In multi-input multi-output control, there are some methods to order reduction of systems such as Truncation and Residualization. Is it applicable for unstable MIMO systems ?
Model reduction should be applied to table plants only. For SISO systems you can factorize the system into stable and unstable parts and perform the model reduction only on the stable part. For MIMO systems, you can use Hankel singular value reduction in MATLAB by using "reduce" command in MATLAB.
- Badges
- Science topic