Model Predictive control -CARIMA Model (Discrete transfer functions)

Hello,

I have started study on Model predictive control (MPC). For time being im using CARIMA model to predict outputs only for discrete T.F(Z-domain).I have implemented it in MATLAB and had obtain the required Matrices (H,P & Q). Now I'm not being able to interpret/understand how to see the response of the new predicted system,using these matrices. It would be really kind of you guys.
Regards,
M. Hamza
 

Attachments

Hello

Thanks for sharing !

Here one video that can highlight some points that you are tryin to reach out!


Here the vid about step response:


I am pretty that can be a good support for better understanding on that MPC with Carima.

Hope this can help,
Regards,
James
 
Well what is the studied system ...
May we know what you is the core subject ...

We may know to guide you

James
I have started my research in MPC.New to it.I'm unable to find a good basic litrature on this topic with full implementation.The best i found are this professor john rossiter's videos.He explains very good right from the start.But now i'm having some difficulty in using these obtained matrices to generate a graphical response to see and interpret the results.I'm unable to create uk and del uk+1 vectors .
 
I have started my research in MPC.New to it.I'm unable to find a good basic litrature on this topic with full implementation.The best i found are this professor john rossiter's videos.He explains very good right from the start.But now i'm having some difficulty in using these obtained matrices to generate a graphical response to see and interpret the results.I'm unable to create uk and del uk+1 vectors .
Well what is the studied system ...
May we know what you is the core subject ...

We may know to guide you

James
Thanks alot for your time. mush appreciated.
 
I don't have Matlab so I don't know what that error means. Look at the link I sent. You may be better off using Python and the gecko package to get started. You really do need a good model to start with.
 
Top