CODEWITHSUNDEEP
modelicadymola
Jack
Jack

Reputation: 1232

how to do multi-loop PID Tuning in the Modelica model?

Tuning the PID parameter is crucial in the control system development, here is an example of how to do this for just one PID controller. https://www.claytex.com/tech-blog/tuning-pid-gains-dymola-modelling/

But if there are multiple PID controllers in the Modelica model, IS there a standard procedure of tuning the parameters?

Is there a Modelica library for multiple PID controller tuning?

enter image description here

Upvotes: 0

Views: 376

Answers (1)

Markus A.
Markus A.

Reputation: 7500

Is there a standard procedure to tuning parameters: Likely not, as this a pretty big area in engineering. It strongly depends on the type of system you need to control, although you limited to PID already, for which a set of tuning methods exist.

In the Dymola Manual 2 there is a section about tuning parameters using the Design library based on the example Design.Optimization.Examples.ControllerDesign_F14. The tutorial related is given in section "First optimization setup". It uses optimization and should be rather generally applicable. Also the LinearSystems2 library can be of great help, as it provides tools for controller design, e.g. a state-space representation of a model (after linearization).

Regarding your specific problem, it is a bit difficult to judge from your screenshot, but assuming it is cascaded control, this should help. In case the controlled variables influence each other they should be decoupled, as e.g. shown here. If non of the before is the case and there are no strong limitations expected in the actuators manual tuning seems feasible.

BTW: Modelica is especially well suited for complex control problems as it can use the plant model to do inverse control. But this seems to be a bit over-engineered for your application.

Upvotes: 5

Related Questions